☆ Efficient Induction of Language Models Via Probabilistic Concept Formation

Christopher J. MacLellan, Peter Matsakis, Pat Langley

This paper presents a novel approach to the acquisition of language models from corpora. The framework builds on Cobweb, an early system for constructing taxonomic hierarchies of probabilistic concepts that used a tabular, attribute-value encoding of training cases and concepts, making it unsuitable for sequential input like language. In response, we explore three new extensions to Cobweb -- the Word, Leaf, and Path variants. These systems encode each training case as an anchor word and surrounding context words, and they store probabilistic descriptions of concepts as distributions over anchor and context information. As in the original Cobweb, a performance element sorts a new instance downward through the hierarchy and uses the final node to predict missing features. Learning is interleaved with performance, updating concept probabilities and hierarchy structure as classification occurs. Thus, the new approaches process training cases in an incremental, online manner that it very different from most methods for statistical language learning. We examine how well the three variants place synonyms together and keep homonyms apart, their ability to recall synonyms as a function of training set size, and their training efficiency. Finally, we discuss related work on incremental learning and directions for further research.

comment: 18 pages, 5 figures, Presented at Advances in Cognitive Systems 2022

☆ Multilingual News Location Detection using an Entity-Based Siamese Network with Semi-Supervised Contrastive Learning and Knowledge Base

Víctor Suárez-Paniagua, Steven Derby, Tri Kurniawan Wijaya

Early detection of relevant locations in a piece of news is especially important in extreme events such as environmental disasters, war conflicts, disease outbreaks, or political turmoils. Additionally, this detection also helps recommender systems to promote relevant news based on user locations. Note that, when the relevant locations are not mentioned explicitly in the text, state-of-the-art methods typically fail to recognize them because these methods rely on syntactic recognition. In contrast, by incorporating a knowledge base and connecting entities with their locations, our system successfully infers the relevant locations even when they are not mentioned explicitly in the text. To evaluate the effectiveness of our approach, and due to the lack of datasets in this area, we also contribute to the research community with a gold-standard multilingual news-location dataset, NewsLOC. It contains the annotation of the relevant locations (and their WikiData IDs) of 600+ Wikinews articles in five different languages: English, French, German, Italian, and Spanish. Through experimental evaluations, we show that our proposed system outperforms the baselines and the fine-tuned version of the model using semi-supervised data that increases the classification rate. The source code and the NewsLOC dataset are publicly available for being used by the research community at https://github.com/vsuarezpaniagua/NewsLocation.

☆ Training Integer-Only Deep Recurrent Neural Networks

Vahid Partovi Nia, Eyyüb Sari, Vanessa Courville, Masoud Asgharian

Recurrent neural networks (RNN) are the backbone of many text and speech applications. These architectures are typically made up of several computationally complex components such as; non-linear activation functions, normalization, bi-directional dependence and attention. In order to maintain good accuracy, these components are frequently run using full-precision floating-point computation, making them slow, inefficient and difficult to deploy on edge devices. In addition, the complex nature of these operations makes them challenging to quantize using standard quantization methods without a significant performance drop. We present a quantization-aware training method for obtaining a highly accurate integer-only recurrent neural network (iRNN). Our approach supports layer normalization, attention, and an adaptive piecewise linear (PWL) approximation of activation functions, to serve a wide range of state-of-the-art RNNs. The proposed method enables RNN-based language models to run on edge devices with $2\times$ improvement in runtime, and $4\times$ reduction in model size while maintaining similar accuracy as its full-precision counterpart.

comment: arXiv admin note: substantial text overlap with arXiv:2109.09828

☆ GENIE: Large Scale Pre-training for Text Generation with Diffusion Model

Zhenghao Lin, Yeyun Gong, Yelong Shen, Tong Wu, Zhihao Fan, Chen Lin, Weizhu Chen, Nan Duan

In this paper, we propose a large-scale language pre-training for text GENeration using dIffusion modEl, which is named GENIE. GENIE is a pre-training sequence-to-sequence text generation model which combines Transformer and diffusion. The diffusion model accepts the latent information from the encoder, which is used to guide the denoising of the current time step. After multiple such denoise iterations, the diffusion model can restore the Gaussian noise to the diverse output text which is controlled by the input text. Moreover, such architecture design also allows us to adopt large scale pre-training on the GENIE. We propose a novel pre-training method named continuous paragraph denoise based on the characteristics of the diffusion model. Extensive experiments on the XSum, CNN/DailyMail, and Gigaword benchmarks shows that GENIE can achieves comparable performance with various strong baselines, especially after pre-training, the generation quality of GENIE is greatly improved. We have also conduct a lot of experiments on the generation diversity and parameter impact of GENIE. The code for GENIE will be made publicly available.

comment: Work in progress

☆ CAMeMBERT: Cascading Assistant-Mediated Multilingual BERT

Dan DeGenaro, Jugal Kalita

Large language models having hundreds of millions, and even billions, of parameters have performed extremely well on a variety of natural language processing (NLP) tasks. Their widespread use and adoption, however, is hindered by the lack of availability and portability of sufficiently large computational resources. This paper proposes a knowledge distillation (KD) technique building on the work of LightMBERT, a student model of multilingual BERT (mBERT). By repeatedly distilling mBERT through increasingly compressed toplayer distilled teacher assistant networks, CAMeMBERT aims to improve upon the time and space complexities of mBERT while keeping loss of accuracy beneath an acceptable threshold. At present, CAMeMBERT has an average accuracy of around 60.1%, which is subject to change after future improvements to the hyperparameters used in fine-tuning.

comment: 4 pages, 2 figures, 3 tables

☆ Understanding Postpartum Parents' Experiences via Two Digital Platforms SC

Xuewen Yao, Miriam Mikhelson, Megan Micheletti, Eunsol Choi, S Craig Watkins, Edison Thomaz, Kaya De Barbaro

Digital platforms, including online forums and helplines, have emerged as avenues of support for caregivers suffering from postpartum mental health distress. Understanding support seekers' experiences as shared on these platforms could provide crucial insight into caregivers' needs during this vulnerable time. In the current work, we provide a descriptive analysis of the concerns, psychological states, and motivations shared by healthy and distressed postpartum support seekers on two digital platforms, a one-on-one digital helpline and a publicly available online forum. Using a combination of human annotations, dictionary models and unsupervised techniques, we find stark differences between the experiences of distressed and healthy mothers. Distressed mothers described interpersonal problems and a lack of support, with 8.60% - 14.56% reporting severe symptoms including suicidal ideation. In contrast, the majority of healthy mothers described childcare issues, such as questions about breastfeeding or sleeping, and reported no severe mental health concerns. Across the two digital platforms, we found that distressed mothers shared similar content. However, the patterns of speech and affect shared by distressed mothers differed between the helpline vs. the online forum, suggesting the design of these platforms may shape meaningful measures of their support-seeking experiences. Our results provide new insight into the experiences of caregivers suffering from postpartum mental health distress. We conclude by discussing methodological considerations for understanding content shared by support seekers and design considerations for the next generation of support tools for postpartum parents.

comment: Will be published in PACM HCI, CSCW1, April 2023 issue

♻ ☆ Chatbots in a Botnet World

Forrest McKee, David Noever

Question-and-answer formats provide a novel experimental platform for investigating cybersecurity questions. Unlike previous chatbots, the latest ChatGPT model from OpenAI supports an advanced understanding of complex coding questions. The research demonstrates thirteen coding tasks that generally qualify as stages in the MITRE ATT&CK framework, ranging from credential access to defense evasion. With varying success, the experimental prompts generate examples of keyloggers, logic bombs, obfuscated worms, and payment-fulfilled ransomware. The empirical results illustrate cases that support the broad gain of functionality, including self-replication and self-modification, evasion, and strategic understanding of complex cybersecurity goals. One surprising feature of ChatGPT as a language-only model centers on its ability to spawn coding approaches that yield images that obfuscate or embed executable programming steps or links.

♻ ☆ Keyphrase Generation with Cross-Document Attention

Shizhe Diao, Yan Song, Tong Zhang

Keyphrase generation aims to produce a set of phrases summarizing the essentials of a given document. Conventional methods normally apply an encoder-decoder architecture to generate the output keyphrases for an input document, where they are designed to focus on each current document so they inevitably omit crucial corpus-level information carried by other similar documents, i.e., the cross-document dependency and latent topics. In this paper, we propose CDKGen, a Transformer-based keyphrase generator, which expands the Transformer to global attention with cross-document attention networks to incorporate available documents as references so as to generate better keyphrases with the guidance of topic information. On top of the proposed Transformer + cross-document attention architecture, we also adopt a copy mechanism to enhance our model via selecting appropriate words from documents to deal with out-of-vocabulary words in keyphrases. Experiment results on five benchmark datasets illustrate the validity and effectiveness of our model, which achieves the state-of-the-art performance on all datasets. Further analyses confirm that the proposed model is able to generate keyphrases consistent with references while keeping sufficient diversity. The code of CDKGen is available at https://github.com/SVAIGBA/CDKGen.

comment: This paper will be superseded by another improved version with new approaches, new settings, and new experimental results

♻ ☆ Towards mapping the contemporary art world with ArtLM: an art-specific NLP model

Qinkai Chen, Mohamed El-Mennaoui, Antoine Fosset, Amine Rebei, Haoyang Cao, Philine Bouscasse, Christy Eóin O'Beirne, Sasha Shevchenko, Mathieu Rosenbaum

With an increasing amount of data in the art world, discovering artists and artworks suitable to collectors' tastes becomes a challenge. It is no longer enough to use visual information, as contextual information about the artist has become just as important in contemporary art. In this work, we present a generic Natural Language Processing framework (called ArtLM) to discover the connections among contemporary artists based on their biographies. In this approach, we first continue to pre-train the existing general English language models with a large amount of unlabelled art-related data. We then fine-tune this new pre-trained model with our biography pair dataset manually annotated by a team of professionals in the art industry. With extensive experiments, we demonstrate that our ArtLM achieves 85.6% accuracy and 84.0% F1 score and outperforms other baseline models. We also provide a visualisation and a qualitative analysis of the artist network built from ArtLM's outputs.

♻ ☆ Compositional generalization in semantic parsing with pretrained transformers

A. Emin Orhan

Large-scale pretraining instills large amounts of knowledge in deep neural networks. This, in turn, improves the generalization behavior of these models in downstream tasks. What exactly are the limits to the generalization benefits of large-scale pretraining? Here, we report observations from some simple experiments aimed at addressing this question in the context of two semantic parsing tasks involving natural language, SCAN and COGS. We show that language models pretrained exclusively with non-English corpora, or even with programming language corpora, significantly improve out-of-distribution generalization in these benchmarks, compared with models trained from scratch, even though both benchmarks are English-based. This demonstrates the surprisingly broad transferability of pretrained representations and knowledge. Pretraining with a large-scale protein sequence prediction task, on the other hand, mostly deteriorates the generalization performance in SCAN and COGS, suggesting that pretrained representations do not transfer universally and that there are constraints on the similarity between the pretraining and downstream domains for successful transfer. Finally, we show that larger models are harder to train from scratch and their generalization accuracy is lower when trained up to convergence on the relatively small SCAN and COGS datasets, but the benefits of large-scale pretraining become much clearer with larger models.

comment: v3 adds one reference, adds further discussion, slightly changes formatting

♻ ☆ SimpleStyle: An Adaptable Style Transfer Approach

Elron Bandel, Yoav Katz, Noam Slonim, Liat Ein-Dor

Attribute-controlled text rewriting, also known as text style-transfer, has a crucial role in regulating attributes and biases of textual training data and a machine generated text. In this work we present SimpleStyle, a minimalist yet effective approach for style-transfer composed of two simple ingredients: controlled denoising and output filtering. Despite the simplicity of our approach, which can be succinctly described with a few lines of code, it is competitive with previous state-of-the-art methods both in automatic and in human evaluation. To demonstrate the adaptability and practical value of our system beyond academic data, we apply SimpleStyle to transfer a wide range of text attributes appearing in real-world textual data from social networks. Additionally, we introduce a novel "soft noising" technique that further improves the performance of our system. We also show that teaching a student model to generate the output of SimpleStyle can result in a system that performs style transfer of equivalent quality with only a single greedy-decoded sample. Finally, we suggest our method as a remedy for the fundamental incompatible baseline issue that holds progress in the field. We offer our protocol as a simple yet strong baseline for works that wish to make incremental advancements in the field of attribute controlled text rewriting.

♻ ☆ NarrativeTime: Dense Temporal Annotation on a Timeline

Anna Rogers, Marzena Karpinska, Ankita Gupta, Vladislav Lialin, Gregory Smelkov, Anna Rumshisky

For the past decade, temporal annotation has been sparse: only a small portion of event pairs in a text was annotated. We present NarrativeTime, the first timeline-based annotation framework that achieves full coverage of all possible TLinks. To compare with the previous SOTA in dense temporal annotation, we perform full re-annotation of TimeBankDense corpus, which shows comparable agreement with a significant increase in density. We contribute TimeBankNT corpus (with each text fully annotated by two expert annotators), extensive annotation guidelines, open-source tools for annotation and conversion to TimeML format, baseline results, as well as quantitative and qualitative analysis of inter-annotator agreement.

♻ ☆ Better Transcription of UK Supreme Court Hearings

Hadeel Saadany, Catherine Breslin, Constantin Orăsan, Sophie Walker

Transcription of legal proceedings is very important to enable access to justice. However, speech transcription is an expensive and slow process. In this paper we describe part of a combined research and industrial project for building an automated transcription tool designed specifically for the Justice sector in the UK. We explain the challenges involved in transcribing court room hearings and the Natural Language Processing (NLP) techniques we employ to tackle these challenges. We will show that fine-tuning a generic off-the-shelf pre-trained Automatic Speech Recognition (ASR) system with an in-domain language model as well as infusing common phrases extracted with a collocation detection model can improve not only the Word Error Rate (WER) of the transcribed hearings but avoid critical errors that are specific of the legal jargon and terminology commonly used in British courts.

♻ ☆ InvBERT: Reconstructing Text from Contextualized Word Embeddings by inverting the BERT pipeline

Kai Kugler, Simon Münker, Johannes Höhmann, Achim Rettinger

Digital Humanities and Computational Literary Studies apply text mining methods to investigate literature. Such automated approaches enable quantitative studies on large corpora which would not be feasible by manual inspection alone. However, due to copyright restrictions, the availability of relevant digitized literary works is limited. Derived Text Formats (DTFs) have been proposed as a solution. Here, textual materials are transformed in such a way that copyright-critical features are removed, but that the use of certain analytical methods remains possible. Contextualized word embeddings produced by transformer-encoders (like BERT) are promising candidates for DTFs because they allow for state-of-the-art performance on various analytical tasks and, at first sight, do not disclose the original text. However, in this paper we demonstrate that under certain conditions the reconstruction of the original copyrighted text becomes feasible and its publication in the form of contextualized token representations is not safe. Our attempts to invert BERT suggest, that publishing the encoder as a black box together with the contextualized embeddings is critical, since it allows to generate data to train a decoder with a reconstruction accuracy sufficient to violate copyright laws.

♻ ☆ Detect-Localize-Repair: A Unified Framework for Learning to Debug with CodeT5 EMNLP 2022

Nghi D. Q. Bui, Yue Wang, Steven Hoi

Automated software debugging is a crucial task for improving the productivity of software developers. Many neural-based techniques have been proven effective for debugging-related tasks such as bug localization and program repair (or bug fixing). However, these techniques often focus only on either one of them or approach them in a stage-wise manner, ignoring the mutual benefits between them. In this work, we propose a novel unified \emph{Detect-Localize-Repair} framework based on a pretrained programming language model CodeT5 to seamlessly address these tasks, named CodeT5-DLR. Specifically, we propose three objectives to adapt the generic CodeT5 for debugging: a bug detection objective to determine whether a given code snippet is buggy or not, a bug localization objective to identify the buggy lines, and a program repair objective to translate the buggy code to its fixed version. We evaluate it on each of these tasks and their combined setting on two newly collected line-level debugging datasets in Java and Python. Extensive results show that our model significantly outperforms existing baselines from both NLP and software engineering domains.

comment: Accepted to EMNLP 2022 Findings Track

♻ ☆ Online Neural Diarization of Unlimited Numbers of Speakers Using Global and Local Attractors

Shota Horiguchi, Shinji Watanabe, Paola Garcia, Yuki Takashima, Yohei Kawaguchi

A method to perform offline and online speaker diarization for an unlimited number of speakers is described in this paper. End-to-end neural diarization (EEND) has achieved overlap-aware speaker diarization by formulating it as a multi-label classification problem. It has also been extended for a flexible number of speakers by introducing speaker-wise attractors. However, the output number of speakers of attractor-based EEND is empirically capped; it cannot deal with cases where the number of speakers appearing during inference is higher than that during training because its speaker counting is trained in a fully supervised manner. Our method, EEND-GLA, solves this problem by introducing unsupervised clustering into attractor-based EEND. In the method, the input audio is first divided into short blocks, then attractor-based diarization is performed for each block, and finally, the results of each block are clustered on the basis of the similarity between locally-calculated attractors. While the number of output speakers is limited within each block, the total number of speakers estimated for the entire input can be higher than the limitation. To use EEND-GLA in an online manner, our method also extends the speaker-tracing buffer, which was originally proposed to enable online inference of conventional EEND. We introduce a block-wise buffer update to make the speaker-tracing buffer compatible with EEND-GLA. Finally, to improve online diarization, our method improves the buffer update method and revisits the variable chunk-size training of EEND. The experimental results demonstrate that EEND-GLA can perform speaker diarization of an unseen number of speakers in both offline and online inferences.

comment: Accepted to IEEE/ACM TASLP

♻ ☆ Localising In-Domain Adaptation of Transformer-Based Biomedical Language Models

Tommaso Mario Buonocore, Claudio Crema, Alberto Redolfi, Riccardo Bellazzi, Enea Parimbelli

In the era of digital healthcare, the huge volumes of textual information generated every day in hospitals constitute an essential but underused asset that could be exploited with task-specific, fine-tuned biomedical language representation models, improving patient care and management. For such specialized domains, previous research has shown that fine-tuning models stemming from broad-coverage checkpoints can largely benefit additional training rounds over large-scale in-domain resources. However, these resources are often unreachable for less-resourced languages like Italian, preventing local medical institutions to employ in-domain adaptation. In order to reduce this gap, our work investigates two accessible approaches to derive biomedical language models in languages other than English, taking Italian as a concrete use-case: one based on neural machine translation of English resources, favoring quantity over quality; the other based on a high-grade, narrow-scoped corpus natively written in Italian, thus preferring quality over quantity. Our study shows that data quantity is a harder constraint than data quality for biomedical adaptation, but the concatenation of high-quality data can improve model performance even when dealing with relatively size-limited corpora. The models published from our investigations have the potential to unlock important research opportunities for Italian hospitals and academia. Finally, the set of lessons learned from the study constitutes valuable insights towards a solution to build biomedical language models that are generalizable to other less-resourced languages and different domain settings.

comment: 8 pages, 3 figures

♻ ☆ An Augmentation Strategy for Visually Rich Documents

Jing Xie, James B. Wendt, Yichao Zhou, Seth Ebner, Sandeep Tata

Many business workflows require extracting important fields from form-like documents (e.g. bank statements, bills of lading, purchase orders, etc.). Recent techniques for automating this task work well only when trained with large datasets. In this work we propose a novel data augmentation technique to improve performance when training data is scarce, e.g. 10-250 documents. Our technique, which we call FieldSwap, works by swapping out the key phrases of a source field with the key phrases of a target field to generate new synthetic examples of the target field for use in training. We demonstrate that this approach can yield 1-7 F1 point improvements in extraction performance.

comment: 9 pages, 6 figures, 3 tables

♻ ☆ Implementation of general formal translators

Iosif Iulian Petrila

The general translator formalism and computing specific implementations are proposed. The implementation of specific elements necessary to process the source and destination information within the translators are presented. Some common directives or instructions, such as classes and procedures, were unified and generalized in order to allow general translations implementations. In order to cover general cases, two levels of processing are required, related to the source and destination information appropriate transformations, with the related control and processing instructions. The proposed general translator elements are useful for processing natural or artificial information described through any types of languages or systems.

♻ ☆ Speech Synthesis with Mixed Emotions

Kun Zhou, Berrak Sisman, Rajib Rana, B. W. Schuller, Haizhou Li

Emotional speech synthesis aims to synthesize human voices with various emotional effects. The current studies are mostly focused on imitating an averaged style belonging to a specific emotion type. In this paper, we seek to generate speech with a mixture of emotions at run-time. We propose a novel formulation that measures the relative difference between the speech samples of different emotions. We then incorporate our formulation into a sequence-to-sequence emotional text-to-speech framework. During the training, the framework does not only explicitly characterize emotion styles, but also explores the ordinal nature of emotions by quantifying the differences with other emotions. At run-time, we control the model to produce the desired emotion mixture by manually defining an emotion attribute vector. The objective and subjective evaluations have validated the effectiveness of the proposed framework. To our best knowledge, this research is the first study on modelling, synthesizing, and evaluating mixed emotions in speech.

comment: Accepted to IEEE Transactions on Affective Computing

♻ ☆ ConvNeXt Based Neural Network for Audio Anti-Spoofing

Qiaowei Ma, Jinghui Zhong, Yitao Yang, Weiheng Liu, Ying Gao, Wing W. Y. Ng

With the rapid development of speech conversion and speech synthesis algorithms, automatic speaker verification (ASV) systems are vulnerable to spoofing attacks. In recent years, researchers had proposed a number of anti-spoofing methods based on hand-crafted features. However, using hand-crafted features rather than raw waveform will lose implicit information for anti-spoofing. Inspired by the promising performance of ConvNeXt in image classification tasks, we revise the ConvNeXt network architecture and propose a lightweight end-to-end anti-spoofing model. By integrating with the channel attention block and using the focal loss function, the proposed model can focus on the most informative sub-bands of speech representations and the difficult samples that are hard to classify. Experiments show that our proposed system could achieve an equal error rate of 0.64% and min-tDCF of 0.0187 for the ASVSpoof 2019 LA evaluation dataset, which outperforms the state-of-the-art systems.

comment: 6 pages

♻ ☆ An Efficient Drug-Drug Interactions Prediction Technology for Molecularly Intelligent Manufacturing

Peng Gao, Feng Gao, Jian-Cheng Ni

Drug-Drug Interactions (DDIs) prediction is an essential issue in the molecular field. Traditional methods of observing DDIs in medical experiments require plenty of resources and labor. In this paper, we present a computational model dubbed MedKGQA based on Graph Neural Networks to automatically predict the DDIs after reading multiple medical documents in the form of multi-hop machine reading comprehension. We introduced a knowledge fusion system to obtain the complete nature of drugs and proteins and exploited a graph reasoning system to infer the drugs and proteins contained in the documents. Our model significantly improves the performance compared to previous state-of-the-art models on the QANGAROO MedHop dataset, which obtained a 4.5% improvement in terms of DDIs prediction accuracy.

☆ DisCoScene: Spatially Disentangled Generative Radiance Fields for Controllable 3D-aware Scene Synthesis

Yinghao Xu, Menglei Chai, Zifan Shi, Sida Peng, Ivan Skorokhodov, Aliaksandr Siarohin, Ceyuan Yang, Yujun Shen, Hsin-Ying Lee, Bolei Zhou, Sergey Tulyakov

Existing 3D-aware image synthesis approaches mainly focus on generating a single canonical object and show limited capacity in composing a complex scene containing a variety of objects. This work presents DisCoScene: a 3Daware generative model for high-quality and controllable scene synthesis. The key ingredient of our method is a very abstract object-level representation (i.e., 3D bounding boxes without semantic annotation) as the scene layout prior, which is simple to obtain, general to describe various scene contents, and yet informative to disentangle objects and background. Moreover, it serves as an intuitive user control for scene editing. Based on such a prior, the proposed model spatially disentangles the whole scene into object-centric generative radiance fields by learning on only 2D images with the global-local discrimination. Our model obtains the generation fidelity and editing flexibility of individual objects while being able to efficiently compose objects and the background into a complete scene. We demonstrate state-of-the-art performance on many scene datasets, including the challenging Waymo outdoor dataset. Project page: https://snap-research.github.io/discoscene/

comment: Project page: https://snap-research.github.io/discoscene/

☆ A Wearable Data Collection System for Studying Micro-Level E-Scooter Behavior in Naturalistic Road Environment

Avinash Prabu, Dan Shen, Renran Tian, Stanley Chien, Lingxi Li, Yaobin Chen, Rini Sherony

As one of the most popular micro-mobility options, e-scooters are spreading in hundreds of big cities and college towns in the US and worldwide. In the meantime, e-scooters are also posing new challenges to traffic safety. In general, e-scooters are suggested to be ridden in bike lanes/sidewalks or share the road with cars at the maximum speed of about 15-20 mph, which is more flexible and much faster than the pedestrains and bicyclists. These features make e-scooters challenging for human drivers, pedestrians, vehicle active safety modules, and self-driving modules to see and interact. To study this new mobility option and address e-scooter riders' and other road users' safety concerns, this paper proposes a wearable data collection system for investigating the micro-level e-Scooter motion behavior in a Naturalistic road environment. An e-Scooter-based data acquisition system has been developed by integrating LiDAR, cameras, and GPS using the robot operating system (ROS). Software frameworks are developed to support hardware interfaces, sensor operation, sensor synchronization, and data saving. The integrated system can collect data continuously for hours, meeting all the requirements including calibration accuracy and capability of collecting the vehicle and e-Scooter encountering data.

comment: Conference: Fast-zero'21, Kanazawa, Japan Date of publication: Sep 2021 Publisher: JSAE

☆ Scalable Adaptive Computation for Iterative Generation

Allan Jabri, David Fleet, Ting Chen

We present the Recurrent Interface Network (RIN), a neural net architecture that allocates computation adaptively to the input according to the distribution of information, allowing it to scale to iterative generation of high-dimensional data. Hidden units of RINs are partitioned into the interface, which is locally connected to inputs, and latents, which are decoupled from inputs and can exchange information globally. The RIN block selectively reads from the interface into latents for high-capacity processing, with incremental updates written back to the interface. Stacking multiple blocks enables effective routing across local and global levels. While routing adds overhead, the cost can be amortized in recurrent computation settings where inputs change gradually while more global context persists, such as iterative generation using diffusion models. To this end, we propose a latent self-conditioning technique that "warm-starts" the latents at each iteration of the generation process. When applied to diffusion models operating directly on pixels, RINs yield state-of-the-art image and video generation without cascades or guidance, while being domain-agnostic and up to 10$\times$ more efficient compared to specialized 2D and 3D U-Nets.

☆ Removing Objects From Neural Radiance Fields

Silvan Weder, Guillermo Garcia-Hernando, Aron Monszpart, Marc Pollefeys, Gabriel Brostow, Michael Firman, Sara Vicente

Neural Radiance Fields (NeRFs) are emerging as a ubiquitous scene representation that allows for novel view synthesis. Increasingly, NeRFs will be shareable with other people. Before sharing a NeRF, though, it might be desirable to remove personal information or unsightly objects. Such removal is not easily achieved with the current NeRF editing frameworks. We propose a framework to remove objects from a NeRF representation created from an RGB-D sequence. Our NeRF inpainting method leverages recent work in 2D image inpainting and is guided by a user-provided mask. Our algorithm is underpinned by a confidence based view selection procedure. It chooses which of the individual 2D inpainted images to use in the creation of the NeRF, so that the resulting inpainted NeRF is 3D consistent. We show that our method for NeRF editing is effective for synthesizing plausible inpaintings in a multi-view coherent manner. We validate our approach using a new and still-challenging dataset for the task of NeRF inpainting.

☆ SupeRGB-D: Zero-shot Instance Segmentation in Cluttered Indoor Environments

Evin Pınar Örnek, Aravindhan K Krishnan, Shreekant Gayaka, Cheng-Hao Kuo, Arnie Sen, Nassir Navab, Federico Tombari

Object instance segmentation is a key challenge for indoor robots navigating cluttered environments with many small objects. Limitations in 3D sensing capabilities often make it difficult to detect every possible object. While deep learning approaches may be effective for this problem, manually annotating 3D data for supervised learning is time-consuming. In this work, we explore zero-shot instance segmentation (ZSIS) from RGB-D data to identify unseen objects in a semantic category-agnostic manner. We introduce a zero-shot split for Tabletop Objects Dataset (TOD-Z) to enable this study and present a method that uses annotated objects to learn the ``objectness'' of pixels and generalize to unseen object categories in cluttered indoor environments. Our method, SupeRGB-D, groups pixels into small patches based on geometric cues and learns to merge the patches in a deep agglomerative clustering fashion. SupeRGB-D outperforms existing baselines on unseen objects while achieving similar performance on seen objects. Additionally, it is extremely lightweight (0.4 MB memory requirement) and suitable for mobile and robotic applications. The dataset split and code will be made publicly available upon acceptance.

☆ Beyond SOT: It's Time to Track Multiple Generic Objects at Once

Christoph Mayer, Martin Danelljan, Ming-Hsuan Yang, Vittorio Ferrari, Luc Van Gool, Alina Kuznetsova

Generic Object Tracking (GOT) is the problem of tracking target objects, specified by bounding boxes in the first frame of a video. While the task has received much attention in the last decades, researchers have almost exclusively focused on the single object setting. Multi-object GOT benefits from a wider applicability, rendering it more attractive in real-world applications. We attribute the lack of research interest into this problem to the absence of suitable benchmarks. In this work, we introduce a new large-scale GOT benchmark, LaGOT, containing multiple annotated target objects per sequence. Our benchmark allows researchers to tackle key remaining challenges in GOT, aiming to increase robustness and reduce computation through joint tracking of multiple objects simultaneously. Furthermore, we propose a Transformer-based GOT tracker TaMOS capable of joint processing of multiple objects through shared computation. TaMOs achieves a 4x faster run-time in case of 10 concurrent objects compared to tracking each object independently and outperforms existing single object trackers on our new benchmark. Finally, TaMOs achieves highly competitive results on single-object GOT datasets, setting a new state-of-the-art on TrackingNet with a success rate AUC of 84.4%. Our benchmark, code, and trained models will be made publicly available.

comment: 16 pages

☆ Fully 3D Implementation of the End-to-end Deep Image Prior-based PET Image Reconstruction Using Block Iterative Algorithm

Fumio Hashimoto, Yuya Onishi, Kibo Ote, Hideaki Tashima, Taiga Yamaya

Deep image prior (DIP) has recently attracted attention owing to its unsupervised positron emission tomography (PET) image reconstruction, which does not require any prior training dataset. In this paper, we present the first attempt to implement an end-to-end DIP-based fully 3D PET image reconstruction method that incorporates a forward-projection model into a loss function. To implement a practical fully 3D PET image reconstruction, which could not be performed due to a graphics processing unit memory limitation, we modify the DIP optimization to block-iteration and sequentially learn an ordered sequence of block sinograms. Furthermore, the relative difference penalty (RDP) term was added to the loss function to enhance the quantitative PET image accuracy. We evaluated our proposed method using Monte Carlo simulation with [$^{18}$F]FDG PET data of a human brain and a preclinical study on monkey brain [$^{18}$F]FDG PET data. The proposed method was compared with the maximum-likelihood expectation maximization (EM), maximum-a-posterior EM with RDP, and hybrid DIP-based PET reconstruction methods. The simulation results showed that the proposed method improved the PET image quality by reducing statistical noise and preserved a contrast of brain structures and inserted tumor compared with other algorithms. In the preclinical experiment, finer structures and better contrast recovery were obtained by the proposed method. This indicated that the proposed method can produce high-quality images without a prior training dataset. Thus, the proposed method is a key enabling technology for the straightforward and practical implementation of end-to-end DIP-based fully 3D PET image reconstruction.

comment: 9 pages, 10 figures

☆ Jamdani Motif Generation using Conditional GAN

MD Tanvir Rouf Shawon, Raihan Tanvir, Humaira Ferdous Shifa, Susmoy Kar, Mohammad Imrul Jubair

Jamdani is the strikingly patterned textile heritage of Bangladesh. The exclusive geometric motifs woven on the fabric are the most attractive part of this craftsmanship having a remarkable influence on textile and fine art. In this paper, we have developed a technique based on the Generative Adversarial Network that can learn to generate entirely new Jamdani patterns from a collection of Jamdani motifs that we assembled, the newly formed motifs can mimic the appearance of the original designs. Users can input the skeleton of a desired pattern in terms of rough strokes and our system finalizes the input by generating the complete motif which follows the geometric structure of real Jamdani ones. To serve this purpose, we collected and preprocessed a dataset containing a large number of Jamdani motifs images from authentic sources via fieldwork and applied a state-of-the-art method called pix2pix to it. To the best of our knowledge, this dataset is currently the only available dataset of Jamdani motifs in digital format for computer vision research. Our experimental results of the pix2pix model on this dataset show satisfactory outputs of computer-generated images of Jamdani motifs and we believe that our work will open a new avenue for further research.

comment: 2020 23rd International Conference on Computer and Information Technology (ICCIT), 2020, pp. 1-6

☆ Monocular 3D Object Detection using Multi-Stage Approaches with Attention and Slicing aided hyper inference

Abonia Sojasingarayar, Ashish Patel

3D object detection is vital as it would enable us to capture objects' sizes, orientation, and position in the world. As a result, we would be able to use this 3D detection in real-world applications such as Augmented Reality (AR), self-driving cars, and robotics which perceive the world the same way we do as humans. Monocular 3D Object Detection is the task to draw 3D bounding box around objects in a single 2D RGB image. It is localization task but without any extra information like depth or other sensors or multiple images. Monocular 3D object detection is an important yet challenging task. Beyond the significant progress in image-based 2D object detection, 3D understanding of real-world objects is an open challenge that has not been explored extensively thus far. In addition to the most closely related studies.

☆ Automatically Annotating Indoor Images with CAD Models via RGB-D Scans

Stefan Ainetter, Sinisa Stekovic, Friedrich Fraundorfer, Vincent Lepetit

We present an automatic method for annotating images of indoor scenes with the CAD models of the objects by relying on RGB-D scans. Through a visual evaluation by 3D experts, we show that our method retrieves annotations that are at least as accurate as manual annotations, and can thus be used as ground truth without the burden of manually annotating 3D data. We do this using an analysis-by-synthesis approach, which compares renderings of the CAD models with the captured scene. We introduce a 'cloning procedure' that identifies objects that have the same geometry, to annotate these objects with the same CAD models. This allows us to obtain complete annotations for the ScanNet dataset and the recent ARKitScenes dataset.

☆ Aliasing is a Driver of Adversarial Attacks

Adrián Rodríguez-Muñoz, Antonio Torralba

Aliasing is a highly important concept in signal processing, as careful consideration of resolution changes is essential in ensuring transmission and processing quality of audio, image, and video. Despite this, up until recently aliasing has received very little consideration in Deep Learning, with all common architectures carelessly sub-sampling without considering aliasing effects. In this work, we investigate the hypothesis that the existence of adversarial perturbations is due in part to aliasing in neural networks. Our ultimate goal is to increase robustness against adversarial attacks using explainable, non-trained, structural changes only, derived from aliasing first principles. Our contributions are the following. First, we establish a sufficient condition for no aliasing for general image transformations. Next, we study sources of aliasing in common neural network layers, and derive simple modifications from first principles to eliminate or reduce it. Lastly, our experimental results show a solid link between anti-aliasing and adversarial attacks. Simply reducing aliasing already results in more robust classifiers, and combining anti-aliasing with robust training out-performs solo robust training on $L_2$ attacks with none or minimal losses in performance on $L_{\infty}$ attacks.

comment: 14 pages, 9 figures, 4 tables

☆ Deep Simplex Classifier for Maximizing the Margin in Both Euclidean and Angular Spaces

Hakan Cevikalp, Hasan Saribas

The classification loss functions used in deep neural network classifiers can be grouped into two categories based on maximizing the margin in either Euclidean or angular spaces. Euclidean distances between sample vectors are used during classification for the methods maximizing the margin in Euclidean spaces whereas the Cosine similarity distance is used during the testing stage for the methods maximizing margin in the angular spaces. This paper introduces a novel classification loss that maximizes the margin in both the Euclidean and angular spaces at the same time. This way, the Euclidean and Cosine distances will produce similar and consistent results and complement each other, which will in turn improve the accuracies. The proposed loss function enforces the samples of classes to cluster around the centers that represent them. The centers approximating classes are chosen from the boundary of a hypersphere, and the pairwise distances between class centers are always equivalent. This restriction corresponds to choosing centers from the vertices of a regular simplex. There is not any hyperparameter that must be set by the user in the proposed loss function, therefore the use of the proposed method is extremely easy for classical classification problems. Moreover, since the class samples are compactly clustered around their corresponding means, the proposed classifier is also very suitable for open set recognition problems where test samples can come from the unknown classes that are not seen in the training phase. Experimental studies show that the proposed method achieves the state-of-the-art accuracies on open set recognition despite its simplicity.

☆ Depth Estimation maps of lidar and stereo images

Fei Wu, Luoyu Chen

This paper as technology report is focusing on evaluation and performance about depth estimations based on lidar data and stereo images(front left and front right). The lidar 3d cloud data and stereo images are provided by ford. In addition, this paper also will explain some details about optimization for depth estimation performance. And some reasons why not use machine learning to do depth estimation, replaced by pure mathmatics to do stereo depth estimation. The structure of this paper is made of by following:(1) Performance: to discuss and evaluate about depth maps created from stereo images and 3D cloud points, and relationships analysis for alignment and errors;(2) Depth estimation by stereo images: to explain the methods about how to use stereo images to estimate depth;(3)Depth estimation by lidar: to explain the methods about how to use 3d cloud datas to estimate depth;In summary, this report is mainly to show the performance of depth maps and their approaches, analysis for them.

comment: 10 pages, 13 figures

☆ GOOD: Exploring Geometric Cues for Detecting Objects in an Open World ICLR 2023

Haiwen Huang, Andreas Geiger, Dan Zhang

We address the task of open-world class-agnostic object detection, i.e., detecting every object in an image by learning from a limited number of base object classes. State-of-the-art RGB-based models suffer from overfitting the training classes and often fail at detecting novel-looking objects. This is because RGB-based models primarily rely on appearance similarity to detect novel objects and are also prone to overfitting short-cut cues such as textures and discriminative parts. To address these shortcomings of RGB-based object detectors, we propose incorporating geometric cues such as depth and normals, predicted by general-purpose monocular estimators. Specifically, we use the geometric cues to train an object proposal network for pseudo-labeling unannotated novel objects in the training set. Our resulting Geometry-guided Open-world Object Detector (GOOD) significantly improves detection recall for novel object categories and already performs well with only a few training classes. Using a single "person" class for training on the COCO dataset, GOOD surpasses SOTA methods by 5.0% AR@100, a relative improvement of 24%.

comment: Under review as a conference paper at ICLR 2023

☆ Reversible Column Networks

Yuxuan Cai, Yizhuang Zhou, Qi Han, Jianjian Sun, Xiangwen Kong, Jun Li, Xiangyu Zhang

We propose a new neural network design paradigm Reversible Column Network (RevCol). The main body of RevCol is composed of multiple copies of subnetworks, named columns respectively, between which multi-level reversible connections are employed. Such architectural scheme attributes RevCol very different behavior from conventional networks: during forward propagation, features in RevCol are learned to be gradually disentangled when passing through each column, whose total information is maintained rather than compressed or discarded as other network does. Our experiments suggest that CNN-style RevCol models can achieve very competitive performances on multiple computer vision tasks such as image classification, object detection and semantic segmentation, especially with large parameter budget and large dataset. For example, after ImageNet-22K pre-training, RevCol-XL obtains 88.2% ImageNet-1K accuracy. Given more pre-training data, our largest model RevCol-H reaches 90.0% on ImageNet-1K, 63.8% APbox on COCO detection minival set, 61.0% mIoU on ADE20k segmentation. To our knowledge, it is the best COCO detection and ADE20k segmentation result among pure (static) CNN models. Moreover, as a general macro architecture fashion, RevCol can also be introduced into transformers or other neural networks, which is demonstrated to improve the performances in both computer vision and NLP tasks. We release code and models at https://github.com/megvii-research/RevCol

☆ Timestamp-Supervised Action Segmentation in the Perspective of Clustering

Dazhao Du, Enhan Li, Lingyu Si, Fanjiang Xu, Fuchun Sun

Video action segmentation aims to slice the video into several action segments. Recently, timestamp supervision has received much attention due to lower annotation costs. We find the frames near the boundaries of action segments are in the transition region between two consecutive actions and have unclear semantics, which we call ambiguous intervals. Most existing methods iteratively generate pseudo-labels for all frames in each video to train the segmentation model. However, ambiguous intervals are more likely to be assigned with noisy and incorrect pseudo-labels, which leads to performance degradation. We propose a novel framework to train the model under timestamp supervision including the following two parts. First, pseudo-label ensembling generates pseudo-label sequences with ambiguous intervals, where the frames have no pseudo-labels. Second, iterative clustering iteratively propagates the pseudo-labels to the ambiguous intervals by clustering, and thus updates the pseudo-label sequences to train the model. We further introduce a clustering loss, which encourages the features of frames within the same action segment more compact. Extensive experiments show the effectiveness of our method.

☆ Predictive Coding Based Multiscale Network with Encoder-Decoder LSTM for Video Prediction

Chaofan Ling, Junpei Zhong, Weihua Li

We are introducing a multi-scale predictive model for video prediction here, whose design is inspired by the "Predictive Coding" theories and "Coarse to Fine" approach. As a predictive coding model, it is updated by a combination of bottom-up and top-down information flows, which is different from traditional bottom-up training style. Its advantage is to reduce the dependence on input information and improve its ability to predict and generate images. Importantly, we achieve with a multi-scale approach -- higher level neurons generate coarser predictions (lower resolution), while the lower level generate finer predictions (higher resolution). This is different from the traditional predictive coding framework in which higher level predict the activity of neurons in lower level. To improve the predictive ability, we integrate an encoder-decoder network in the LSTM architecture and share the final encoded high-level semantic information between different levels. Additionally, since the output of each network level is an RGB image, a smaller LSTM hidden state can be used to retain and update the only necessary hidden information, avoiding being mapped to an overly discrete and complex space. In this way, we can reduce the difficulty of prediction and the computational overhead. Finally, we further explore the training strategies, to address the instability in adversarial training and mismatch between training and testing in long-term prediction. Code is available at https://github.com/Ling-CF/MSPN.

☆ Hybrid Quantum-Classical Generative Adversarial Network for High Resolution Image Generation

Shu Lok Tsang, Maxwell T. West, Sarah M. Erfani, Muhammad Usman

Quantum machine learning (QML) has received increasing attention due to its potential to outperform classical machine learning methods in various problems. A subclass of QML methods is quantum generative adversarial networks (QGANs) which have been studied as a quantum counterpart of classical GANs widely used in image manipulation and generation tasks. The existing work on QGANs is still limited to small-scale proof-of-concept examples based on images with significant down-scaling. Here we integrate classical and quantum techniques to propose a new hybrid quantum-classical GAN framework. We demonstrate its superior learning capabilities by generating $28 \times 28$ pixels grey-scale images without dimensionality reduction or classical pre/post-processing on multiple classes of the standard MNIST and Fashion MNIST datasets, which achieves comparable results to classical frameworks with 3 orders of magnitude less trainable generator parameters. To gain further insight into the working of our hybrid approach, we systematically explore the impact of its parameter space by varying the number of qubits, the size of image patches, the number of layers in the generator, the shape of the patches and the choice of prior distribution. Our results show that increasing the quantum generator size generally improves the learning capability of the network. The developed framework provides a foundation for future design of QGANs with optimal parameter set tailored for complex image generation tasks.

☆ DDColor: Towards Photo-Realistic and Semantic-Aware Image Colorization via Dual Decoders

Xiaoyang Kang, Tao Yang, Wenqi Ouyang, Peiran Ren, Lingzhi Li, Xuansong Xie

Automatic image colorization is a particularly challenging problem. Due to the high illness of the problem and multi-modal uncertainty, directly training a deep neural network usually leads to incorrect semantic colors and low color richness. Existing transformer-based methods can deliver better results but highly depend on hand-crafted dataset-level empirical distribution priors. In this work, we propose DDColor, a new end-to-end method with dual decoders, for image colorization. More specifically, we design a multi-scale image decoder and a transformer-based color decoder. The former manages to restore the spatial resolution of the image, while the latter establishes the correlation between semantic representations and color queries via cross-attention. The two decoders incorporate to learn semantic-aware color embedding by leveraging the multi-scale visual features. With the help of these two decoders, our method succeeds in producing semantically consistent and visually plausible colorization results without any additional priors. In addition, a simple but effective colorfulness loss is introduced to further improve the color richness of generated results. Our extensive experiments demonstrate that the proposed DDColor achieves significantly superior performance to existing state-of-the-art works both quantitatively and qualitatively. Codes will be made publicly available.

☆ Deformable Surface Reconstruction via Riemannian Metric Preservation

Oriol Barbany, Adrià Colomé, Carme Torras

Estimating the pose of an object from a monocular image is an inverse problem fundamental in computer vision. The ill-posed nature of this problem requires incorporating deformation priors to solve it. In practice, many materials do not perceptibly shrink or extend when manipulated, constituting a powerful and well-known prior. Mathematically, this translates to the preservation of the Riemannian metric. Neural networks offer the perfect playground to solve the surface reconstruction problem as they can approximate surfaces with arbitrary precision and allow the computation of differential geometry quantities. This paper presents an approach to inferring continuous deformable surfaces from a sequence of images, which is benchmarked against several techniques and obtains state-of-the-art performance without the need for offline training.

☆ Metadata-guided Consistency Learning for High Content Images

Johan Fredin Haslum, Christos Matsoukas, Karl-Johan Leuchowius, Erik Müllers, Kevin Smith

High content imaging assays can capture rich phenotypic response data for large sets of compound treatments, aiding in the characterization and discovery of novel drugs. However, extracting representative features from high content images that can capture subtle nuances in phenotypes remains challenging. The lack of high-quality labels makes it difficult to achieve satisfactory results with supervised deep learning. Self-Supervised learning methods, which learn from automatically generated labels has shown great success on natural images, offer an attractive alternative also to microscopy images. However, we find that self-supervised learning techniques underperform on high content imaging assays. One challenge is the undesirable domain shifts present in the data known as batch effects, which may be caused by biological noise or uncontrolled experimental conditions. To this end, we introduce Cross-Domain Consistency Learning (CDCL), a novel approach that is able to learn in the presence of batch effects. CDCL enforces the learning of biological similarities while disregarding undesirable batch-specific signals, which leads to more useful and versatile representations. These features are organised according to their morphological changes and are more useful for downstream tasks - such as distinguishing treatments and mode of action.

☆ Tune-A-Video: One-Shot Tuning of Image Diffusion Models for Text-to-Video Generation

Jay Zhangjie Wu, Yixiao Ge, Xintao Wang, Weixian Lei, Yuchao Gu, Wynne Hsu, Ying Shan, Xiaohu Qie, Mike Zheng Shou

To reproduce the success of text-to-image (T2I) generation, recent works in text-to-video (T2V) generation employ large-scale text-video dataset for fine-tuning. However, such paradigm is computationally expensive. Humans have the amazing ability to learn new visual concepts from just one single exemplar. We hereby study a new T2V generation problem$\unicode{x2014}$One-Shot Video Generation, where only a single text-video pair is presented for training an open-domain T2V generator. Intuitively, we propose to adapt the T2I diffusion model pretrained on massive image data for T2V generation. We make two key observations: 1) T2I models are able to generate images that align well with the verb terms; 2) extending T2I models to generate multiple images concurrently exhibits surprisingly good content consistency. To further learn continuous motion, we propose Tune-A-Video with a tailored Sparse-Causal Attention, which generates videos from text prompts via an efficient one-shot tuning of pretrained T2I diffusion models. Tune-A-Video is capable of producing temporally-coherent videos over various applications such as change of subject or background, attribute editing, style transfer, demonstrating the versatility and effectiveness of our method.

comment: Preprint

☆ DaDe: Delay-adoptive Detector for Streaming Perception

Wonwoo Jo, Kyungshin Lee, Jaewon Baik, Sangsun Lee, Dongho Choi, Hyunkyoo Park

Recognizing the surrounding environment at low latency is critical in autonomous driving. In real-time environment, surrounding environment changes when processing is over. Current detection models are incapable of dealing with changes in the environment that occur after processing. Streaming perception is proposed to assess the latency and accuracy of real-time video perception. However, additional problems arise in real-world applications due to limited hardware resources, high temperatures, and other factors. In this study, we develop a model that can reflect processing delays in real time and produce the most reasonable results. By incorporating the proposed feature queue and feature select module, the system gains the ability to forecast specific time steps without any additional computational costs. Our method is tested on the Argoverse-HD dataset. It achieves higher performance than the current state-of-the-art methods(2022.10) in various environments when delayed . The code is available at https://github.com/danjos95/DADE

comment: This paper is accepted in VISAPP - 2023 (Full Paper, Oral Presentation)

☆ Ultra-High-Definition Low-Light Image Enhancement: A Benchmark and Transformer-Based Method AAAI 2023

Tao Wang, Kaihao Zhang, Tianrun Shen, Wenhan Luo, Bjorn Stenger, Tong Lu

As the quality of optical sensors improves, there is a need for processing large-scale images. In particular, the ability of devices to capture ultra-high definition (UHD) images and video places new demands on the image processing pipeline. In this paper, we consider the task of low-light image enhancement (LLIE) and introduce a large-scale database consisting of images at 4K and 8K resolution. We conduct systematic benchmarking studies and provide a comparison of current LLIE algorithms. As a second contribution, we introduce LLFormer, a transformer-based low-light enhancement method. The core components of LLFormer are the axis-based multi-head self-attention and cross-layer attention fusion block, which significantly reduces the linear complexity. Extensive experiments on the new dataset and existing public datasets show that LLFormer outperforms state-of-the-art methods. We also show that employing existing LLIE methods trained on our benchmark as a pre-processing step significantly improves the performance of downstream tasks, e.g., face detection in low-light conditions. The source code and pre-trained models are available at https://github.com/TaoWangzj/LLFormer.

comment: Accepted at AAAI 2023. #AAAI2023

☆ Mask Focal Loss for dense crowd counting with canonical object detection networks

Xiaopin Zhong, Guankun Wang, Weixiang Liua, Zongze Wua, Yuanlong Deng

Crowd counting plays an important role in risk perception and early warning, traffic control and scene statistical analysis. The challenges of crowd counting in highly dense and complex scenes lie in the mutual occlusion of the human body parts, the large variation of the body scales and the complexity of imaging conditions. Deep learning based head detection is a promising method for crowd counting. However the highly concerned object detection networks cannot be well applied to this field for two main reasons. First, most of the existing head detection datasets are only annotated with the center points instead of bounding boxes which is mandatory for the canonical detectors. Second, the sample imbalance has not been overcome yet in highly dense and complex scenes because the existing loss functions calculate the positive loss at a single key point or in the entire target area with the same weight. To address these problems, We propose a novel loss function, called Mask Focal Loss, to unify the loss functions based on heatmap ground truth (GT) and binary feature map GT. Mask Focal Loss redefines the weight of the loss contributions according to the situ value of the heatmap with a Gaussian kernel. For better evaluation and comparison, a new synthetic dataset GTA\_Head is made public, including 35 sequences, 5096 images and 1732043 head labels with bounding boxes. Experimental results show the overwhelming performance and demonstrate that our proposed Mask Focal Loss is applicable to all of the canonical detectors and to various datasets with different GT. This provides a strong basis for surpassing the crowd counting methods based on density estimation.

☆ Generative Colorization of Structured Mobile Web Pages WACV 2023

Kotaro Kikuchi, Naoto Inoue, Mayu Otani, Edgar Simo-Serra, Kota Yamaguchi

Color is a critical design factor for web pages, affecting important factors such as viewer emotions and the overall trust and satisfaction of a website. Effective coloring requires design knowledge and expertise, but if this process could be automated through data-driven modeling, efficient exploration and alternative workflows would be possible. However, this direction remains underexplored due to the lack of a formalization of the web page colorization problem, datasets, and evaluation protocols. In this work, we propose a new dataset consisting of e-commerce mobile web pages in a tractable format, which are created by simplifying the pages and extracting canonical color styles with a common web browser. The web page colorization problem is then formalized as a task of estimating plausible color styles for a given web page content with a given hierarchical structure of the elements. We present several Transformer-based methods that are adapted to this task by prepending structural message passing to capture hierarchical relationships between elements. Experimental results, including a quantitative evaluation designed for this task, demonstrate the advantages of our methods over statistical and image colorization methods. The code is available at https://github.com/CyberAgentAILab/webcolor.

comment: Accepted to WACV 2023

☆ SHLE: Devices Tracking and Depth Filtering for Stereo-based Height Limit Estimation

Zhaoxin Fan, Kaixing Yang, Min Zhang, Zhenbo Song, Hongyan Liu, Jun He

Recently, over-height vehicle strike frequently occurs, causing great economic cost and serious safety problems. Hence, an alert system which can accurately discover any possible height limiting devices in advance is necessary to be employed in modern large or medium sized cars, such as touring cars. Detecting and estimating the height limiting devices act as the key point of a successful height limit alert system. Though there are some works research height limit estimation, existing methods are either too computational expensive or not accurate enough. In this paper, we propose a novel stereo-based pipeline named SHLE for height limit estimation. Our SHLE pipeline consists of two stages. In stage 1, a novel devices detection and tracking scheme is introduced, which accurately locate the height limit devices in the left or right image. Then, in stage 2, the depth is temporally measured, extracted and filtered to calculate the height limit device. To benchmark the height limit estimation task, we build a large-scale dataset named "Disparity Height", where stereo images, pre-computed disparities and ground-truth height limit annotations are provided. We conducted extensive experiments on "Disparity Height" and the results show that SHLE achieves an average error below than 10cm though the car is 70m away from the devices. Our method also outperforms all compared baselines and achieves state-of-the-art performance. Code is available at https://github.com/Yang-Kaixing/SHLE.

☆ LaneAF: Robust Multi-Lane Detection with Affinity Fields

Genze Zhou, Luoyu Chen, Fei Wu

Lane detection is a long-standing task and a basic module in autonomous driving. The task is to detect the lane of the current driving road, and provide relevant information such as the ID, direction, curvature, width, length, with visualization. Our work is based on CNN backbone DLA-34, along with Affinity Fields, aims to achieve robust detection of various lanes without assuming the number of lanes. Besides, we investigate novel decoding methods to achieve more efficient lane detection algorithm.

comment: 6 pages, 7 figures

☆ Confidence-Aware Paced-Curriculum Learning by Label Smoothing for Surgical Scene Understanding

Mengya Xu, Mobarakol Islam, Ben Glocker, Hongliang Ren

Curriculum learning and self-paced learning are the training strategies that gradually feed the samples from easy to more complex. They have captivated increasing attention due to their excellent performance in robotic vision. Most recent works focus on designing curricula based on difficulty levels in input samples or smoothing the feature maps. However, smoothing labels to control the learning utility in a curriculum manner is still unexplored. In this work, we design a paced curriculum by label smoothing (P-CBLS) using paced learning with uniform label smoothing (ULS) for classification tasks and fuse uniform and spatially varying label smoothing (SVLS) for semantic segmentation tasks in a curriculum manner. In ULS and SVLS, a bigger smoothing factor value enforces a heavy smoothing penalty in the true label and limits learning less information. Therefore, we design the curriculum by label smoothing (CBLS). We set a bigger smoothing value at the beginning of training and gradually decreased it to zero to control the model learning utility from lower to higher. We also designed a confidence-aware pacing function and combined it with our CBLS to investigate the benefits of various curricula. The proposed techniques are validated on four robotic surgery datasets of multi-class, multi-label classification, captioning, and segmentation tasks. We also investigate the robustness of our method by corrupting validation data into different severity levels. Our extensive analysis shows that the proposed method improves prediction accuracy and robustness.

comment: 12 pages

☆ Supervised Anomaly Detection Method Combining Generative Adversarial Networks and Three-Dimensional Data in Vehicle Inspections

Yohei Baba, Takuro Hoshi, Ryosuke Mori, Gaurang Gavai

The external visual inspections of rolling stock's underfloor equipment are currently being performed via human visual inspection. In this study, we attempt to partly automate visual inspection by investigating anomaly inspection algorithms that use image processing technology. As the railroad maintenance studies tend to have little anomaly data, unsupervised learning methods are usually preferred for anomaly detection; however, training cost and accuracy is still a challenge. Additionally, a researcher created anomalous images from normal images by adding noise, etc., but the anomalous targeted in this study is the rotation of piping cocks that was difficult to create using noise. Therefore, in this study, we propose a new method that uses style conversion via generative adversarial networks on three-dimensional computer graphics and imitates anomaly images to apply anomaly detection based on supervised learning. The geometry-consistent style conversion model was used to convert the image, and because of this the color and texture of the image were successfully made to imitate the real image while maintaining the anomalous shape. Using the generated anomaly images as supervised data, the anomaly detection model can be easily trained without complex adjustments and successfully detects anomalies.

comment: 6 pages, 12 figures

☆ Group Sparse Coding for Image Denoising

Luoyu Chen, Fei Wu

Group sparse representation has shown promising results in image debulrring and image inpainting in GSR [3] , the main reason that lead to the success is by exploiting Sparsity and Nonlocal self-similarity (NSS) between patches on natural images, and solve a regularized optimization problem. However, directly adapting GSR[3] in image denoising yield very unstable and non-satisfactory results, to overcome these issues, this paper proposes a progressive image denoising algorithm that successfully adapt GSR [3] model and experiments shows the superior performance than some of the state-of-the-art methods.

comment: 6 pages, 5 figures

☆ Understanding and Improving the Role of Projection Head in Self-Supervised Learning

Kartik Gupta, Thalaiyasingam Ajanthan, Anton van den Hengel, Stephen Gould

Self-supervised learning (SSL) aims to produce useful feature representations without access to any human-labeled data annotations. Due to the success of recent SSL methods based on contrastive learning, such as SimCLR, this problem has gained popularity. Most current contrastive learning approaches append a parametrized projection head to the end of some backbone network to optimize the InfoNCE objective and then discard the learned projection head after training. This raises a fundamental question: Why is a learnable projection head required if we are to discard it after training? In this work, we first perform a systematic study on the behavior of SSL training focusing on the role of the projection head layers. By formulating the projection head as a parametric component for the InfoNCE objective rather than a part of the network, we present an alternative optimization scheme for training contrastive learning based SSL frameworks. Our experimental study on multiple image classification datasets demonstrates the effectiveness of the proposed approach over alternatives in the SSL literature.

☆ SALVE: Self-supervised Adaptive Low-light Video Enhancement

Zohreh Azizi, C. -C. Jay Kuo

A self-supervised adaptive low-light video enhancement (SALVE) method is proposed in this work. SALVE first conducts an effective Retinex-based low-light image enhancement on a few key frames of an input low-light video. Next, it learns mappings from the low- to enhanced-light frames via Ridge regression. Finally, it uses these mappings to enhance the remaining frames in the input video. SALVE is a hybrid method that combines components from a traditional Retinex-based image enhancement method and a learning-based method. The former component leads to a robust solution which is easily adaptive to new real-world environments. The latter component offers a fast, computationally inexpensive and temporally consistent solution. We conduct extensive experiments to show the superior performance of SALVE. Our user study shows that 87% of participants prefer SALVE over prior work.

comment: 11 pages, 7 figures, 2 tables

☆ Spatio-Visual Fusion-Based Person Re-Identification for Overhead Fisheye Images

Mertcan Cokbas, Prakash Ishwar, Janusz Konrad

Reliable and cost-effective counting of people in large indoor spaces is a significant challenge with many applications. An emerging approach is to deploy multiple fisheye cameras mounted overhead to monitor the whole space. However, due to the overlapping fields of view, person re-identificaiton (PRID) is critical for the accuracy of counting. While PRID has been thoroughly researched for traditional rectilinear cameras, few methods have been proposed for fisheye cameras and their performance is comparatively lower. To close this performance gap, we propose a multi-feature framework for fisheye PRID where we combine deep-learning, color-based and location-based features by means of novel feature fusion. We evaluate the performance of our framework for various feature combinations on FRIDA, a public fisheye PRID dataset. The results demonstrate that our multi-feature approach outperforms recent appearance-based deep-learning methods by almost 18% points and location-based methods by almost 3% points in accuracy.

☆ Restoring Vision in Hazy Weather with Hierarchical Contrastive Learning

Tao Wang, Guangpin Tao, Wanglong Lu, Kaihao Zhang, Wenhan Luo, Xiaoqin Zhang, Tong Lu

Image restoration under hazy weather condition, which is called single image dehazing, has been of significant interest for various computer vision applications. In recent years, deep learning-based methods have achieved success. However, existing image dehazing methods typically neglect the hierarchy of features in the neural network and fail to exploit their relationships fully. To this end, we propose an effective image dehazing method named Hierarchical Contrastive Dehazing (HCD), which is based on feature fusion and contrastive learning strategies. HCD consists of a hierarchical dehazing network (HDN) and a novel hierarchical contrastive loss (HCL). Specifically, the core design in the HDN is a Hierarchical Interaction Module, which utilizes multi-scale activation to revise the feature responses hierarchically. To cooperate with the training of HDN, we propose HCL which performs contrastive learning on hierarchically paired exemplars, facilitating haze removal. Extensive experiments on public datasets, RESIDE, HazeRD, and DENSE-HAZE, demonstrate that HCD quantitatively outperforms the state-of-the-art methods in terms of PSNR, SSIM and achieves better visual quality.

comment: 27 pages, 9 figures

☆ Multi-queue Momentum Contrast for Microvideo-Product Retrieval WSDM '23

Yali Du, Yinwei Wei, Wei Ji, Fan Liu, Xin Luo, Liqiang Nie

The booming development and huge market of micro-videos bring new e-commerce channels for merchants. Currently, more micro-video publishers prefer to embed relevant ads into their micro-videos, which not only provides them with business income but helps the audiences to discover their interesting products. However, due to the micro-video recording by unprofessional equipment, involving various topics and including multiple modalities, it is challenging to locate the products related to micro-videos efficiently, appropriately, and accurately. We formulate the microvideo-product retrieval task, which is the first attempt to explore the retrieval between the multi-modal and multi-modal instances. A novel approach named Multi-Queue Momentum Contrast (MQMC) network is proposed for bidirectional retrieval, consisting of the uni-modal feature and multi-modal instance representation learning. Moreover, a discriminative selection strategy with a multi-queue is used to distinguish the importance of different negatives based on their categories. We collect two large-scale microvideo-product datasets (MVS and MVS-large) for evaluation and manually construct the hierarchical category ontology, which covers sundry products in daily life. Extensive experiments show that MQMC outperforms the state-of-the-art baselines. Our replication package (including code, dataset, etc.) is publicly available at https://github.com/duyali2000/MQMC.

comment: Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining (WSDM '23), February 27-March 3, 2023, Singapore, Singapore

☆ IPProtect: protecting the intellectual property of visual datasets during data valuation

Gursimran Singh, Chendi Wang, Ahnaf Tazwar, Lanjun Wang, Yong Zhang

Data trading is essential to accelerate the development of data-driven machine learning pipelines. The central problem in data trading is to estimate the utility of a seller's dataset with respect to a given buyer's machine learning task, also known as data valuation. Typically, data valuation requires one or more participants to share their raw dataset with others, leading to potential risks of intellectual property (IP) violations. In this paper, we tackle the novel task of preemptively protecting the IP of datasets that need to be shared during data valuation. First, we identify and formalize two kinds of novel IP risks in visual datasets: data-item (image) IP and statistical (dataset) IP. Then, we propose a novel algorithm to convert the raw dataset into a sanitized version, that provides resistance to IP violations, while at the same time allowing accurate data valuation. The key idea is to limit the transfer of information from the raw dataset to the sanitized dataset, thereby protecting against potential intellectual property violations. Next, we analyze our method for the likely existence of a solution and immunity against reconstruction attacks. Finally, we conduct extensive experiments on three computer vision datasets demonstrating the advantages of our method in comparison to other baselines.

comment: 14 pages, 5 figures

☆ Vision-Based Environmental Perception for Autonomous Driving

Fei Liu, Zihao Lu, Xianke Lin

Visual perception plays an important role in autonomous driving. One of the primary tasks is object detection and identification. Since the vision sensor is rich in color and texture information, it can quickly and accurately identify various road information. The commonly used technique is based on extracting and calculating various features of the image. The recent development of deep learning-based method has better reliability and processing speed and has a greater advantage in recognizing complex elements. For depth estimation, vision sensor is also used for ranging due to their small size and low cost. Monocular camera uses image data from a single viewpoint as input to estimate object depth. In contrast, stereo vision is based on parallax and matching feature points of different views, and the application of deep learning also further improves the accuracy. In addition, Simultaneous Location and Mapping (SLAM) can establish a model of the road environment, thus helping the vehicle perceive the surrounding environment and complete the tasks. In this paper, we introduce and compare various methods of object detection and identification, then explain the development of depth estimation and compare various methods based on monocular, stereo, and RDBG sensors, next review and compare various methods of SLAM, and finally summarize the current problems and present the future development trends of vision technologies.

comment: 39 pages, 17 figures

☆ Offline Clustering Approach to Self-supervised Learning for Class-imbalanced Image Data

Hye-min Chang, Sungkyun Chang

Class-imbalanced datasets are known to cause the problem of model being biased towards the majority classes. In this project, we set up two research questions: 1) when is the class-imbalance problem more prevalent in self-supervised pre-training? and 2) can offline clustering of feature representations help pre-training on class-imbalanced data? Our experiments investigate the former question by adjusting the degree of {\it class-imbalance} when training the baseline models, namely SimCLR and SimSiam on CIFAR-10 database. To answer the latter question, we train each expert model on each subset of the feature clusters. We then distill the knowledge of expert models into a single model, so that we will be able to compare the performance of this model to our baselines.

comment: 5 pages, 3 figures, Technical report

☆ Novel Deep Learning Framework For Bovine Iris Segmentation

Heemoon Yoon, Mira Park, Sang-Hee Lee

Iris segmentation is the initial step to identify biometric of animals to establish a traceability system of livestock. In this study, we propose a novel deep learning framework for pixel-wise segmentation with minimum use of annotation labels using BovineAAEyes80 public dataset. In the experiment, U-Net with VGG16 backbone was selected as the best combination of encoder and decoder model, demonstrating a 99.50% accuracy and a 98.35% Dice coefficient score. Remarkably, the selected model accurately segmented corrupted images even without proper annotation data. This study contributes to the advancement of the iris segmentation and the development of a reliable DNNs training framework.

comment: 5 pages, 4 figures, 3 tables

♻ ☆ Edge2Vec: A High Quality Embedding for the Jigsaw Puzzle Problem

Daniel Rika, Dror Sholomon, Eli David, Nathan S. Netanyahu

Pairwise compatibility measure (CM) is a key component in solving the jigsaw puzzle problem (JPP) and many of its recently proposed variants. With the rapid rise of deep neural networks (DNNs), a trade-off between performance (i.e., accuracy) and computational efficiency has become a very significant issue. Whereas an end-to-end DNN-based CM model exhibits high performance, it becomes virtually infeasible on very large puzzles, due to its highly intensive computation. On the other hand, exploiting the concept of embeddings to alleviate significantly the computational efficiency, has resulted in degraded performance, according to recent studies. This paper derives an advanced CM model (based on modified embeddings and a new loss function, called hard batch triplet loss) for closing the above gap between speed and accuracy; namely a CM model that achieves SOTA results in terms of performance and efficiency combined. We evaluated our newly derived CM on three commonly used datasets, and obtained a reconstruction improvement of 5.8% and 19.5% for so-called Type-1 and Type-2 problem variants, respectively, compared to best known results due to previous CMs.

♻ ☆ MetaFormer Baselines for Vision

Weihao Yu, Chenyang Si, Pan Zhou, Mi Luo, Yichen Zhou, Jiashi Feng, Shuicheng Yan, Xinchao Wang

MetaFormer, the abstracted architecture of Transformer, has been found to play a significant role in achieving competitive performance. In this paper, we further explore the capacity of MetaFormer, again, without focusing on token mixer design: we introduce several baseline models under MetaFormer using the most basic or common mixers, and summarize our observations as follows: (1) MetaFormer ensures solid lower bound of performance. By merely adopting identity mapping as the token mixer, the MetaFormer model, termed IdentityFormer, achieves >80% accuracy on ImageNet-1K. (2) MetaFormer works well with arbitrary token mixers. When specifying the token mixer as even a random matrix to mix tokens, the resulting model RandFormer yields an accuracy of >81%, outperforming IdentityFormer. Rest assured of MetaFormer's results when new token mixers are adopted. (3) MetaFormer effortlessly offers state-of-the-art results. With just conventional token mixers dated back five years ago, the models instantiated from MetaFormer already beat state of the art. (a) ConvFormer outperforms ConvNeXt. Taking the common depthwise separable convolutions as the token mixer, the model termed ConvFormer, which can be regarded as pure CNNs, outperforms the strong CNN model ConvNeXt. (b) CAFormer sets new record on ImageNet-1K. By simply applying depthwise separable convolutions as token mixer in the bottom stages and vanilla self-attention in the top stages, the resulting model CAFormer sets a new record on ImageNet-1K: it achieves an accuracy of 85.5% at 224x224 resolution, under normal supervised training without external data or distillation. In our expedition to probe MetaFormer, we also find that a new activation, StarReLU, reduces 71% FLOPs of activation compared with GELU yet achieves better performance. We expect StarReLU to find great potential in MetaFormer-like models alongside other neural networks.

comment: Add more ImageNet-22K pretrained models. Code: https://github.com/sail-sg/metaformer

♻ ☆ EDICT: Exact Diffusion Inversion via Coupled Transformations

Bram Wallace, Akash Gokul, Nikhil Naik

Finding an initial noise vector that produces an input image when fed into the diffusion process (known as inversion) is an important problem in denoising diffusion models (DDMs), with applications for real image editing. The state-of-the-art approach for real image editing with inversion uses denoising diffusion implicit models (DDIMs) to deterministically noise the image to the intermediate state along the path that the denoising would follow given the original conditioning. However, DDIM inversion for real images is unstable as it relies on local linearization assumptions, which result in the propagation of errors, leading to incorrect image reconstruction and loss of content. To alleviate these problems, we propose Exact Diffusion Inversion via Coupled Transformations (EDICT), an inversion method that draws inspiration from affine coupling layers. EDICT enables mathematically exact inversion of real and model-generated images by maintaining two coupled noise vectors which are used to invert each other in an alternating fashion. Using Stable Diffusion, a state-of-the-art latent diffusion model, we demonstrate that EDICT successfully reconstructs real images with high fidelity. On complex image datasets like MS-COCO, EDICT reconstruction significantly outperforms DDIM, improving the mean square error of reconstruction by a factor of two. Using noise vectors inverted from real images, EDICT enables a wide range of image edits--from local and global semantic edits to image stylization--while maintaining fidelity to the original image structure. EDICT requires no model training/finetuning, prompt tuning, or extra data and can be combined with any pretrained DDM. Code is available at https://github.com/salesforce/EDICT.

comment: 24 pages, 22 figures. Code now available

♻ ☆ Unlocking the potential of two-point cells for energy-efficient and resilient training of deep nets

Ahsan Adeel, Adewale Adetomi, Khubaib Ahmed, Amir Hussain, Tughrul Arslan, W. A. Phillips

Context-sensitive two-point layer 5 pyramidal cells (L5PCs) were discovered as long ago as 1999. However, the potential of this discovery to provide useful neural computation has yet to be demonstrated. Here we show for the first time how a transformative L5PCs-driven deep neural network (DNN), termed the multisensory cooperative computing (MCC) architecture, can effectively process large amounts of heterogeneous real-world audio-visual (AV) data, using far less energy compared to best available 'point' neuron-driven DNNs. A novel highly-distributed parallel implementation on a Xilinx UltraScale+ MPSoC device estimates energy savings up to 245759 $ \times $ 50000 $\mu$J (i.e., 62% less than the baseline model in a semi-supervised learning setup) where a single synapse consumes $8e^{-5}\mu$J. In a supervised learning setup, the energy-saving can potentially reach up to 1250x less (per feedforward transmission) than the baseline model. The significantly reduced neural activity in MCC leads to inherently fast learning and resilience against sudden neural damage. This remarkable performance in pilot experiments demonstrates the embodied neuromorphic intelligence of our proposed cooperative L5PC that receives input from diverse neighbouring neurons as context to amplify the transmission of most salient and relevant information for onward transmission, from overwhelmingly large multimodal information utilised at the early stages of on-chip training. Our proposed approach opens new cross-disciplinary avenues for future on-chip DNN training implementations and posits a radical shift in current neuromorphic computing paradigms.

♻ ☆ EgoEnv: Human-centric environment representations from egocentric video

Tushar Nagarajan, Santhosh Kumar Ramakrishnan, Ruta Desai, James Hillis, Kristen Grauman

First-person video highlights a camera-wearer's activities in the context of their persistent environment. However, current video understanding approaches reason over visual features from short video clips that are detached from the underlying physical space and capture only what is immediately visible. We present an approach that links egocentric video and the environment by learning representations that are predictive of the camera-wearer's (potentially unseen) local surroundings to facilitate human-centric environment understanding. We train such models using videos from agents in simulated 3D environments where the environment is fully observable, and test them on human-captured real-world videos from unseen environments. On two human-centric video tasks, we show that state-of-the-art video models equipped with our environment-aware features consistently outperform their counterparts with traditional clip features. Moreover, despite being trained exclusively on simulated videos, our approach successfully handles real-world videos from HouseTours and Ego4D. Project page: https://vision.cs.utexas.edu/projects/ego-env/

♻ ☆ Multi-task Learning for Camera Calibration

Talha Hanif Butt, Murtaza Taj

For a number of tasks, such as 3D reconstruction, robotic interface, autonomous driving, etc., camera calibration is essential. In this study, we present a unique method for predicting intrinsic (principal point offset and focal length) and extrinsic (baseline, pitch, and translation) properties from a pair of images. We suggested a novel method where camera model equations are represented as a neural network in a multi-task learning framework, in contrast to existing methods, which build a comprehensive solution. By reconstructing the 3D points using a camera model neural network and then using the loss in reconstruction to obtain the camera specifications, this innovative camera projection loss (CPL) method allows us that the desired parameters should be estimated. As far as we are aware, our approach is the first one that uses an approach to multi-task learning that includes mathematical formulas in a framework for learning to estimate camera parameters to predict both the extrinsic and intrinsic parameters jointly. Additionally, we provided a new dataset named as CVGL Camera Calibration Dataset [1] which has been collected using the CARLA Simulator [2]. Actually, we show that our suggested strategy out performs both conventional methods and methods based on deep learning on 8 out of 10 parameters that were assessed using both real and synthetic data. Our code and generated dataset are available at https://github.com/thanif/Camera-Calibration-through-Camera-Projection-Loss.

comment: 20 pages, 12 figures, 8 tables

♻ ☆ ReViSe: Remote Vital Signs Measurement Using Smartphone Camera

Donghao Qiao, Amtul Haq Ayesha, Farhana Zulkernine, Raihan Masroor, Nauman Jaffar

We propose an end-to-end framework to measure people's vital signs including Heart Rate (HR), Heart Rate Variability (HRV), Oxygen Saturation (SpO2) and Blood Pressure (BP) based on the rPPG methodology from the video of a user's face captured with a smartphone camera. We extract face landmarks with a deep learning-based neural network model in real-time. Multiple face patches also called Regions-of-Interest (RoIs) are extracted by using the predicted face landmarks. Several filters are applied to reduce the noise from the RoIs in the extracted cardiac signals called Blood Volume Pulse (BVP) signal. The measurements of HR, HRV and SpO2 are validated on two public rPPG datasets namely the TokyoTech rPPG and the Pulse Rate Detection (PURE) datasets, on which our models achieved the following Mean Absolute Errors (MAE): a) for HR, 1.73Beats-Per-Minute (bpm) and 3.95bpm respectively; b) for HRV, 18.55ms and 25.03ms respectively, and c) for SpO2, an MAE of 1.64% on the PURE dataset. We validated our end-to-end rPPG framework, ReViSe, in daily living environment, and thereby created the Video-HR dataset. Our HR estimation model achieved an MAE of 2.49bpm on this dataset. Since no publicly available rPPG datasets existed for BP measurement with face videos, we used a dataset with signals from fingertip sensor to train our deep learning-based BP estimation model and also created our own video dataset, Video-BP. On our Video-BP dataset, our BP estimation model achieved an MAE of 6.7mmHg for Systolic Blood Pressure (SBP), and an MAE of 9.6mmHg for Diastolic Blood Pressure (DBP). ReViSe framework has been validated on datasets with videos recorded in daily living environment as opposed to less noisy laboratory environment as reported by most state-of-the-art techniques.

♻ ☆ Contrastive learning of Class-agnostic Activation Map for Weakly Supervised Object Localization and Semantic Segmentation CVPR 2022

Jinheng Xie, Jianfeng Xiang, Junliang Chen, Xianxu Hou, Xiaodong Zhao, Linlin Shen

While class activation map (CAM) generated by image classification network has been widely used for weakly supervised object localization (WSOL) and semantic segmentation (WSSS), such classifiers usually focus on discriminative object regions. In this paper, we propose Contrastive learning for Class-agnostic Activation Map (C$^2$AM) generation only using unlabeled image data, without the involvement of image-level supervision. The core idea comes from the observation that i) semantic information of foreground objects usually differs from their backgrounds; ii) foreground objects with similar appearance or background with similar color/texture have similar representations in the feature space. We form the positive and negative pairs based on the above relations and force the network to disentangle foreground and background with a class-agnostic activation map using a novel contrastive loss. As the network is guided to discriminate cross-image foreground-background, the class-agnostic activation maps learned by our approach generate more complete object regions. We successfully extracted from C$^2$AM class-agnostic object bounding boxes for object localization and background cues to refine CAM generated by classification network for semantic segmentation. Extensive experiments on CUB-200-2011, ImageNet-1K, and PASCAL VOC2012 datasets show that both WSOL and WSSS can benefit from the proposed C$^2$AM.

comment: Accepted by CVPR 2022

♻ ☆ A3GC-IP: Attention-Oriented Adjacency Adaptive Recurrent Graph Convolutions for Human Pose Estimation from Sparse Inertial Measurements

Patrik Puchert, Timo Ropinski

Conventional methods for human pose estimation either require a high degree of instrumentation, by relying on many inertial measurement units (IMUs), or constraint the recording space, by relying on extrinsic cameras. These deficits are tackled through the approach of human pose estimation from sparse IMU data. We define attention-oriented adjacency adaptive graph convolutional long-short term memory networks (A3GC-LSTM), to tackle human pose estimation based on six IMUs, through incorporating the human body graph structure directly into the network. The A3GC-LSTM combines both spatial and temporal dependency in a single network operation, more memory efficiently than previous approaches. The recurrent graph learning on arbitrarily long sequences is made possible by equipping graph convolutions with adjacency adaptivity, which eliminates the problem of information loss in deep or recurrent graph networks, while it also allows for learning unknown dependencies between the human body joints. To further boost accuracy, a spatial attention formalism is incorporated into the recurrent LSTM cell. With our presented approach, we are able to utilize the inherent graph nature of the human body, and thus can outperform the state of the art for human pose estimation from sparse IMU data.

comment: Preprint, in submission

♻ ☆ Graph Matching with Bi-level Noisy Correspondence

Yijie Lin, Mouxing Yang, Jun Yu, Peng Hu, Changqing Zhang, Xi Peng

In this paper, we study a novel and widely existing problem in graph matching (GM), namely, Bi-level Noisy Correspondence (BNC), which refers to node-level noisy correspondence (NNC) and edge-level noisy correspondence (ENC). In brief, on the one hand, due to the poor recognizability and viewpoint differences between images, it is inevitable to inaccurately annotate some keypoints with offset and confusion, leading to the mismatch between two associated nodes, i.e., NNC. On the other hand, the noisy node-to-node correspondence will further contaminate the edge-to-edge correspondence, thus leading to ENC. For the BNC challenge, we propose a novel method termed Contrastive Matching with Momentum Distillation. Specifically, the proposed method is with a robust quadratic contrastive loss which enjoys the following merits: i) better exploring the node-to-node and edge-to-edge correlations through a GM customized quadratic contrastive learning paradigm; ii) adaptively penalizing the noisy assignments based on the confidence estimated by the momentum teacher. Extensive experiments on three real-world datasets show the robustness of our model compared with 12 competitive baselines.

♻ ☆ Sparsely-gated MoE Layers for CNN Interpretability

Svetlana Pavlitskaya, Christian Hubschneider, Lukas Struppek, J. Marius Zöllner

Sparsely-gated Mixture of Expert (MoE) layers have been recently successfully applied for scaling large transformers, especially for language modeling tasks. An intriguing side effect of sparse MoE layers is that they convey inherent interpretability to a model via natural expert specialization. In this work, we apply sparse MoE layers to CNNs for computer vision tasks and analyze the resulting effect on model interpretability. To stabilize MoE training, we present both soft and hard constraint-based approaches. With hard constraints, the weights of certain experts are allowed to become zero, while soft constraints balance the contribution of experts with an additional auxiliary loss. As a result, soft constraints handle expert utilization better and support the expert specialization process, while hard constraints maintain more generalized experts and increase overall model performance. Our findings demonstrate that experts can implicitly focus on individual sub-domains of the input space. For example, experts trained for CIFAR-100 image classification specialize in recognizing different domains such as flowers or animals without previous data clustering. Experiments with RetinaNet and the COCO dataset further indicate that object detection experts can also specialize in detecting objects of distinct sizes.

♻ ☆ Localizing Anatomical Landmarks in Ocular Images using Zoom-In Attentive Networks

Xiaofeng Lei, Shaohua Li, Xinxing Xu, Huazhu Fu, Yong Liu, Yih-Chung Tham, Yangqin Feng, Mingrui Tan, Yanyu Xu, Jocelyn Hui Lin Goh, Rick Siow Mong Goh, Ching-Yu Cheng

Localizing anatomical landmarks are important tasks in medical image analysis. However, the landmarks to be localized often lack prominent visual features. Their locations are elusive and easily confused with the background, and thus precise localization highly depends on the context formed by their surrounding areas. In addition, the required precision is usually higher than segmentation and object detection tasks. Therefore, localization has its unique challenges different from segmentation or detection. In this paper, we propose a zoom-in attentive network (ZIAN) for anatomical landmark localization in ocular images. First, a coarse-to-fine, or "zoom-in" strategy is utilized to learn the contextualized features in different scales. Then, an attentive fusion module is adopted to aggregate multi-scale features, which consists of 1) a co-attention network with a multiple regions-of-interest (ROIs) scheme that learns complementary features from the multiple ROIs, 2) an attention-based fusion module which integrates the multi-ROIs features and non-ROI features. We evaluated ZIAN on two open challenge tasks, i.e., the fovea localization in fundus images and scleral spur localization in AS-OCT images. Experiments show that ZIAN achieves promising performances and outperforms state-of-the-art localization methods. The source code and trained models of ZIAN are available at https://github.com/leixiaofeng-astar/OMIA9-ZIAN.

♻ ☆ Suppress with a Patch: Revisiting Universal Adversarial Patch Attacks against Object Detection CEC

Svetlana Pavlitskaya, Jonas Hendl, Sebastian Kleim, Leopold Müller, Fabian Wylczoch, J. Marius Zöllner

Adversarial patch-based attacks aim to fool a neural network with an intentionally generated noise, which is concentrated in a particular region of an input image. In this work, we perform an in-depth analysis of different patch generation parameters, including initialization, patch size, and especially positioning a patch in an image during training. We focus on the object vanishing attack and run experiments with YOLOv3 as a model under attack in a white-box setting and use images from the COCO dataset. Our experiments have shown, that inserting a patch inside a window of increasing size during training leads to a significant increase in attack strength compared to a fixed position. The best results were obtained when a patch was positioned randomly during training, while patch position additionally varied within a batch.

comment: Accepted for publication at ICECCME 2022

♻ ☆ Task-adaptive Spatial-Temporal Video Sampler for Few-shot Action Recognition ACM MM 2022

Huabin Liu, Weixian Lv, John See, Weiyao Lin

A primary challenge faced in few-shot action recognition is inadequate video data for training. To address this issue, current methods in this field mainly focus on devising algorithms at the feature level while little attention is paid to processing input video data. Moreover, existing frame sampling strategies may omit critical action information in temporal and spatial dimensions, which further impacts video utilization efficiency. In this paper, we propose a novel video frame sampler for few-shot action recognition to address this issue, where task-specific spatial-temporal frame sampling is achieved via a temporal selector (TS) and a spatial amplifier (SA). Specifically, our sampler first scans the whole video at a small computational cost to obtain a global perception of video frames. The TS plays its role in selecting top-T frames that contribute most significantly and subsequently. The SA emphasizes the discriminative information of each frame by amplifying critical regions with the guidance of saliency maps. We further adopt task-adaptive learning to dynamically adjust the sampling strategy according to the episode task at hand. Both the implementations of TS and SA are differentiable for end-to-end optimization, facilitating seamless integration of our proposed sampler with most few-shot action recognition methods. Extensive experiments show a significant boost in the performances on various benchmarks including long-term videos.The code is available at https://github.com/R00Kie-Liu/Sampler

comment: Accepted by ACM MM 2022

♻ ☆ TA2N: Two-Stage Action Alignment Network for Few-shot Action Recognition AAAI 2022

Shuyuan Li, Huabin Liu, Rui Qian, Yuxi Li, John See, Mengjuan Fei, Xiaoyuan Yu, Weiyao Lin

Few-shot action recognition aims to recognize novel action classes (query) using just a few samples (support). The majority of current approaches follow the metric learning paradigm, which learns to compare the similarity between videos. Recently, it has been observed that directly measuring this similarity is not ideal since different action instances may show distinctive temporal distribution, resulting in severe misalignment issues across query and support videos. In this paper, we arrest this problem from two distinct aspects -- action duration misalignment and action evolution misalignment. We address them sequentially through a Two-stage Action Alignment Network (TA2N). The first stage locates the action by learning a temporal affine transform, which warps each video feature to its action duration while dismissing the action-irrelevant feature (e.g. background). Next, the second stage coordinates query feature to match the spatial-temporal action evolution of support by performing temporally rearrange and spatially offset prediction. Extensive experiments on benchmark datasets show the potential of the proposed method in achieving state-of-the-art performance for few-shot action recognition.The code of this project can be found at https://github.com/R00Kie-Liu/TA2N

comment: Published in AAAI 2022

♻ ☆ Position Embedding Needs an Independent Layer Normalization

Runyi Yu, Zhennan Wang, Yinhuai Wang, Kehan Li, Yian Zhao, Jian Zhang, Guoli Song, Jie Chen

The Position Embedding (PE) is critical for Vision Transformers (VTs) due to the permutation-invariance of self-attention operation. By analyzing the input and output of each encoder layer in VTs using reparameterization and visualization, we find that the default PE joining method (simply adding the PE and patch embedding together) operates the same affine transformation to token embedding and PE, which limits the expressiveness of PE and hence constrains the performance of VTs. To overcome this limitation, we propose a simple, effective, and robust method. Specifically, we provide two independent layer normalizations for token embeddings and PE for each layer, and add them together as the input of each layer's Muti-Head Self-Attention module. Since the method allows the model to adaptively adjust the information of PE for different layers, we name it as Layer-adaptive Position Embedding, abbreviated as LaPE. Extensive experiments demonstrate that LaPE can improve various VTs with different types of PE and make VTs robust to PE types. For example, LaPE improves 0.94% accuracy for ViT-Lite on Cifar10, 0.98% for CCT on Cifar100, and 1.72% for DeiT on ImageNet-1K, which is remarkable considering the negligible extra parameters, memory and computational cost brought by LaPE. The code is publicly available at https://github.com/Ingrid725/LaPE.

comment: 14 pages, 8 figures

♻ ☆ PS-NeRF: Neural Inverse Rendering for Multi-view Photometric Stereo ECCV 2022

Wenqi Yang, Guanying Chen, Chaofeng Chen, Zhenfang Chen, Kwan-Yee K. Wong

Traditional multi-view photometric stereo (MVPS) methods are often composed of multiple disjoint stages, resulting in noticeable accumulated errors. In this paper, we present a neural inverse rendering method for MVPS based on implicit representation. Given multi-view images of a non-Lambertian object illuminated by multiple unknown directional lights, our method jointly estimates the geometry, materials, and lights. Our method first employs multi-light images to estimate per-view surface normal maps, which are used to regularize the normals derived from the neural radiance field. It then jointly optimizes the surface normals, spatially-varying BRDFs, and lights based on a shadow-aware differentiable rendering layer. After optimization, the reconstructed object can be used for novel-view rendering, relighting, and material editing. Experiments on both synthetic and real datasets demonstrate that our method achieves far more accurate shape reconstruction than existing MVPS and neural rendering methods. Our code and model can be found at https://ywq.github.io/psnerf.

comment: ECCV 2022, Project page: https://ywq.github.io/psnerf

♻ ☆ Flame-state monitoring based on very low number of visible or infrared images via few-shot learning

Ruiyuan Kang, Panos Liatsis, Dimitrios C. Kyritsis

The current success of machine learning on image-based combustion monitoring is based on massive data, which is costly even impossible for industrial applications. To address this conflict, we introduce few-shot learning in order to achieve combustion monitoring and classification for the first time. Two algorithms, Siamese Network coupled with k Nearest Neighbors (SN-kNN) and Prototypical Network (PN), were tested. Rather than utilizing solely visible images as discussed in previous studies, we also used Infrared (IR) images. We analyzed the training process, test performance and inference speed of two algorithms on both image formats, and also used t-SNE to visualize learned features. The results demonstrated that both SN-kNN and PN were capable to distinguish flame states from learning with merely 20 images per flame state. The worst performance, which was realized by PN on IR images, still possessed precision, accuracy, recall, and F1-score above 0.95. We showed that visible images demonstrated more substantial differences between classes and presented more consistent patterns inside the class, which made the training speed and model performance better compared to IR images. In contrast, the relatively low quality of IR images made it difficult for PN to extract distinguishable prototypes, which caused relatively weak performance. With the entrire training set supporting classification, SN-kNN performed well with IR images. On the other hand, benefitting from the architecture design, PN has a much faster speed in training and inference than SN-kNN. The presented work analyzed the characteristics of both algorithms and image formats for the first time, thus providing guidance for their future utilization in combustion monitoring tasks.

comment: 16 pages, 12 figures, four tables

♻ ☆ Dataset Condensation with Distribution Matching

Bo Zhao, Hakan Bilen

Computational cost of training state-of-the-art deep models in many learning problems is rapidly increasing due to more sophisticated models and larger datasets. A recent promising direction for reducing training cost is dataset condensation that aims to replace the original large training set with a significantly smaller learned synthetic set while preserving the original information. While training deep models on the small set of condensed images can be extremely fast, their synthesis remains computationally expensive due to the complex bi-level optimization and second-order derivative computation. In this work, we propose a simple yet effective method that synthesizes condensed images by matching feature distributions of the synthetic and original training images in many sampled embedding spaces. Our method significantly reduces the synthesis cost while achieving comparable or better performance. Thanks to its efficiency, we apply our method to more realistic and larger datasets with sophisticated neural architectures and obtain a significant performance boost. We also show promising practical benefits of our method in continual learning and neural architecture search.

♻ ☆ Semi-Supervised Learning of Monocular Depth Estimation via Consistency Regularization with K-way Disjoint Masking

Jongbeom Baek, Gyeongnyeon Kim, Seonghoon Park, Honggyu An, Matteo Poggi, Seungryong Kim

Semi-Supervised Learning (SSL) has recently accomplished successful achievements in various fields such as image classification, object detection, and semantic segmentation, which typically require a lot of labour to construct ground-truth. Especially in the depth estimation task, annotating training data is very costly and time-consuming, and thus recent SSL regime seems an attractive solution. In this paper, for the first time, we introduce a novel framework for semi-supervised learning of monocular depth estimation networks, using consistency regularization to mitigate the reliance on large ground-truth depth data. We propose a novel data augmentation approach, called K-way disjoint masking, which allows the network for learning how to reconstruct invisible regions so that the model not only becomes robust to perturbations but also generates globally consistent output depth maps. Experiments on the KITTI and NYU-Depth-v2 datasets demonstrate the effectiveness of each component in our pipeline, robustness to the use of fewer and fewer annotated images, and superior results compared to other state-of-the-art, semi-supervised methods for monocular depth estimation. Our code is available at https://github.com/KU-CVLAB/MaskingDepth.

comment: Project page: https://github.com/KU-CVLAB/MaskingDepth

♻ ☆ DenseGAP: Graph-Structured Dense Correspondence Learning with Anchor Points

Zhengfei Kuang, Jiaman Li, Mingming He, Tong Wang, Yajie Zhao

Establishing dense correspondence between two images is a fundamental computer vision problem, which is typically tackled by matching local feature descriptors. However, without global awareness, such local features are often insufficient for disambiguating similar regions. And computing the pairwise feature correlation across images is both computation-expensive and memory-intensive. To make the local features aware of the global context and improve their matching accuracy, we introduce DenseGAP, a new solution for efficient Dense correspondence learning with a Graph-structured neural network conditioned on Anchor Points. Specifically, we first propose a graph structure that utilizes anchor points to provide sparse but reliable prior on inter- and intra-image context and propagates them to all image points via directed edges. We also design a graph-structured network to broadcast multi-level contexts via light-weighted message-passing layers and generate high-resolution feature maps at low memory cost. Finally, based on the predicted feature maps, we introduce a coarse-to-fine framework for accurate correspondence prediction using cycle consistency. Our feature descriptors capture both local and global information, thus enabling a continuous feature field for querying arbitrary points at high resolution. Through comprehensive ablative experiments and evaluations on large-scale indoor and outdoor datasets, we demonstrate that our method advances the state-of-the-art of correspondence learning on most benchmarks.

♻ ☆ Self-Supervised Contrastive Representation Learning for 3D Mesh Segmentation AAAI 2023

Ayaan Haque, Hankyu Moon, Heng Hao, Sima Didari, Jae Oh Woo, Patrick Bangert

3D deep learning is a growing field of interest due to the vast amount of information stored in 3D formats. Triangular meshes are an efficient representation for irregular, non-uniform 3D objects. However, meshes are often challenging to annotate due to their high geometrical complexity. Specifically, creating segmentation masks for meshes is tedious and time-consuming. Therefore, it is desirable to train segmentation networks with limited-labeled data. Self-supervised learning (SSL), a form of unsupervised representation learning, is a growing alternative to fully-supervised learning which can decrease the burden of supervision for training. We propose SSL-MeshCNN, a self-supervised contrastive learning method for pre-training CNNs for mesh segmentation. We take inspiration from traditional contrastive learning frameworks to design a novel contrastive learning algorithm specifically for meshes. Our preliminary experiments show promising results in reducing the heavy labeled data requirement needed for mesh segmentation by at least 33%.

comment: AAAI 2023

♻ ☆ DALL-E for Detection: Language-driven Compositional Image Synthesis for Object Detection

Yunhao Ge, Jiashu Xu, Brian Nlong Zhao, Neel Joshi, Laurent Itti, Vibhav Vineet

We propose a new paradigm to automatically generate training data with accurate labels at scale using the text-toimage synthesis frameworks (e.g., DALL-E, Stable Diffusion, etc.). The proposed approach decouples training data generation into foreground object mask generation and background (context) image generation. For foreground object mask generation, we use a simple textual template with object class name as input to DALL-E to generate a diverse set of foreground images. A foreground-background segmentation algorithm is then used to generate foreground object masks. Next, in order to generate context images, first a language description of the context is generated by applying an image captioning method on a small set of images representing the context. These language descriptions are then used to generate diverse sets of context images using the DALL-E framework. These are then composited with object masks generated in the first step to provide an augmented training set for a classifier. We demonstrate the advantages of our approach on four object detection datasets including on Pascal VOC and COCO object detection tasks. Furthermore, we also highlight the compositional nature of our data generation approach on out-of-distribution and zero-shot data generation scenarios.

comment: v3(same as v2) version, update structure (add foreground generation, stable diffusion), add more experiments

☆ Response to Moffat's Comment on "Towards Meaningful Statements in IR Evaluation: Mapping Evaluation Measures to Interval Scales"

Marco Ferrante, Nicola Ferro, Norbert Fuhr

Moffat recently commented on our previous work. Our work focused on how laying the foundations of our evaluation methodology into the theory of measurement can improve our knowledge and understanding of the evaluation measures we use in IR and how it can shed light on the different types of scales adopted by our evaluation measures; we also provided evidence, through extensive experimentation, on the impact of the different types of scales on the statistical analyses, as well as on the impact of departing from their assumptions. Moreover, we investigated, for the first time in IR, the concept of meaningfulness, i.e. the invariance of the experimental statements and inferences you draw, and proposed it as a way to ensure more valid and generalizabile results. Moffat's comments build on: (i) misconceptions about the representational theory of measurement, such as what an interval scale actually is and what axioms it has to comply with; (ii) they totally miss the central concept of meaningfulness. Therefore, we reply to Moffat's comments by properly framing them in the representational theory of measurement and in the concept of meaningfulness. All in all, we can only reiterate what we said several times: the goal of this research line is to theoretically ground our evaluation methodology - and IR is a field where it is extremely challenging to perform any theoretical advances - in order to aim for more robust and generalizable inferences - something we currently lack in the field. Possibly there are other and better ways to achieve this objective and these proposals could emerge from an open discussion in the field and from the work of others. On the other hand, reducing everything to a contrast on what is (or pretend to be) an interval scale or whether all or none evaluation measures are interval scales may be more a barrier from than a help in progressing towards this goal.

☆ Reinforcement Learning Based Approaches to Adaptive Context Caching in Distributed Context Management Systems

Shakthi Weerasinghe, Arkady Zaslavsky, Seng W. Loke, Amin Abken, Alireza Hassani

Performance metrics-driven context caching has a profound impact on throughput and response time in distributed context management systems for real-time context queries. This paper proposes a reinforcement learning based approach to adaptively cache context with the objective of minimizing the cost incurred by context management systems in responding to context queries. Our novel algorithms enable context queries and sub-queries to reuse and repurpose cached context in an efficient manner. This approach is distinctive to traditional data caching approaches by three main features. First, we make selective context cache admissions using no prior knowledge of the context, or the context query load. Secondly, we develop and incorporate innovative heuristic models to calculate expected performance of caching an item when making the decisions. Thirdly, our strategy defines a time-aware continuous cache action space. We present two reinforcement learning agents, a value function estimating actor-critic agent and a policy search agent using deep deterministic policy gradient method. The paper also proposes adaptive policies such as eviction and cache memory scaling to complement our objective. Our method is evaluated using a synthetically generated load of context sub-queries and a synthetic data set inspired from real world data and query samples. We further investigate optimal adaptive caching configurations under different settings. This paper presents, compares, and discusses our findings that the proposed selective caching methods reach short- and long-term cost- and performance-efficiency. The paper demonstrates that the proposed methods outperform other modes of context management such as redirector mode, and database mode, and cache all policy by up to 60% in cost efficiency.

comment: This is a pre-print version of the journal paper submitted to ACM Transactions in Internet of Things, which is currently under review

☆ Local Policy Improvement for Recommender Systems

Dawen Liang, Nikos Vlassis

Recommender systems aim to answer the following question: given the items that a user has interacted with, what items will this user likely interact with next? Historically this problem is often framed as a predictive task via (self-)supervised learning. In recent years, we have seen more emphasis placed on approaching the recommendation problem from a policy optimization perspective: learning a policy that maximizes some reward function (e.g., user engagement). However, it is commonly the case in recommender systems that we are only able to train a new policy given data collected from a previously-deployed policy. The conventional way to address such a policy mismatch is through importance sampling correction, which unfortunately comes with its own limitations. In this paper, we suggest an alternative approach, which involves the use of local policy improvement without off-policy correction. Drawing from a number of related results in the fields of causal inference, bandits, and reinforcement learning, we present a suite of methods that compute and optimize a lower bound of the expected reward of the target policy. Crucially, this lower bound is a function that is easy to estimate from data, and which does not involve density ratios (such as those appearing in importance sampling correction). We argue that this local policy improvement paradigm is particularly well suited for recommender systems, given that in practice the previously-deployed policy is typically of reasonably high quality, and furthermore it tends to be re-trained frequently and gets continuously updated. We discuss some practical recipes on how to apply some of the proposed techniques in a sequential recommendation setting.

♻ ☆ COVID-19 Activity Risk Calculator as a Gamified Public Health Intervention Tool

Shreyasvi Natraj, Malhar Bhide, Nathan Yap, Meng Liu, Agrima Seth, Jonathan Berman, Christin Glorioso

Public health intervention techniques have been highly significant in reducing the negative impact of several epidemics and pandemics. Among all of the wide-spread diseases, one of the most dangerous one has been severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) or Coronavirus disease 2019 (COVID-19). The impact of the virus has been observed in over 200 countries leading to hospitalizations and deaths of millions of people. Currently existing COVID-19 risk estimation tools provided to the general public have been highly variable during the pandemic due to its dependency on rapidly evolving factors such as community transmission levels and variants. There has also been confusion surrounding certain personal protective strategies such as risk reduction by mask-wearing and vaccination. In order to create a simplified easy-to-use tool for estimating different individual risks associated with carrying out daily-life activity, we developed COVID-19 Activity Risk Calculator (CovARC). CovARC serves as a gamified public health intervention as users can "play with" how different risks associated with COVID-19 would change depending on several different factors when carrying out a daily routine activity. Empowering the public to make informed, data-driven decisions about safely engaging in activities may help to reduce COVID- 19 levels in the community. In this study, we demonstrate a streamlined, scalable and accurate COVID-19 risk calculation system. Our study also showcases quantitatively, the increased impact of interventions such as vaccination and mask-wearing when cases are higher, which could prove as a validity to inform and support policy decisions around mask mandate case thresholds and other non-pharmaceutical interventions.

comment: 12 pages, 6 figures (main paper) + 1 figure (supplementary section.)

♻ ☆ Directed Acyclic Graph Factorization Machines for CTR Prediction via Knowledge Distillation

Zhen Tian, Ting Bai, Zibin Zhang, Zhiyuan Xu, Kangyi Lin, Ji-Rong Wen, Wayne Xin Zhao

With the growth of high-dimensional sparse data in web-scale recommender systems, the computational cost to learn high-order feature interaction in CTR prediction task largely increases, which limits the use of high-order interaction models in real industrial applications. Some recent knowledge distillation based methods transfer knowledge from complex teacher models to shallow student models for accelerating the online model inference. However, they suffer from the degradation of model accuracy in knowledge distillation process. It is challenging to balance the efficiency and effectiveness of the shallow student models. To address this problem, we propose a Directed Acyclic Graph Factorization Machine (KD-DAGFM) to learn the high-order feature interactions from existing complex interaction models for CTR prediction via Knowledge Distillation. The proposed lightweight student model DAGFM can learn arbitrary explicit feature interactions from teacher networks, which achieves approximately lossless performance and is proved by a dynamic programming algorithm. Besides, an improved general model KD-DAGFM+ is shown to be effective in distilling both explicit and implicit feature interactions from any complex teacher model. Extensive experiments are conducted on four real-world datasets, including a large-scale industrial dataset from WeChat platform with billions of feature dimensions. KD-DAGFM achieves the best performance with less than 21.5% FLOPs of the state-of-the-art method on both online and offline experiments, showing the superiority of DAGFM to deal with the industrial scale data in CTR prediction task. Our implementation code is available at: https://github.com/RUCAIBox/DAGFM.

☆ Scalable Adaptive Computation for Iterative Generation

Allan Jabri, David Fleet, Ting Chen

We present the Recurrent Interface Network (RIN), a neural net architecture that allocates computation adaptively to the input according to the distribution of information, allowing it to scale to iterative generation of high-dimensional data. Hidden units of RINs are partitioned into the interface, which is locally connected to inputs, and latents, which are decoupled from inputs and can exchange information globally. The RIN block selectively reads from the interface into latents for high-capacity processing, with incremental updates written back to the interface. Stacking multiple blocks enables effective routing across local and global levels. While routing adds overhead, the cost can be amortized in recurrent computation settings where inputs change gradually while more global context persists, such as iterative generation using diffusion models. To this end, we propose a latent self-conditioning technique that "warm-starts" the latents at each iteration of the generation process. When applied to diffusion models operating directly on pixels, RINs yield state-of-the-art image and video generation without cascades or guidance, while being domain-agnostic and up to 10$\times$ more efficient compared to specialized 2D and 3D U-Nets.

☆ Efficient Induction of Language Models Via Probabilistic Concept Formation

Christopher J. MacLellan, Peter Matsakis, Pat Langley

This paper presents a novel approach to the acquisition of language models from corpora. The framework builds on Cobweb, an early system for constructing taxonomic hierarchies of probabilistic concepts that used a tabular, attribute-value encoding of training cases and concepts, making it unsuitable for sequential input like language. In response, we explore three new extensions to Cobweb -- the Word, Leaf, and Path variants. These systems encode each training case as an anchor word and surrounding context words, and they store probabilistic descriptions of concepts as distributions over anchor and context information. As in the original Cobweb, a performance element sorts a new instance downward through the hierarchy and uses the final node to predict missing features. Learning is interleaved with performance, updating concept probabilities and hierarchy structure as classification occurs. Thus, the new approaches process training cases in an incremental, online manner that it very different from most methods for statistical language learning. We examine how well the three variants place synonyms together and keep homonyms apart, their ability to recall synonyms as a function of training set size, and their training efficiency. Finally, we discuss related work on incremental learning and directions for further research.

comment: 18 pages, 5 figures, Presented at Advances in Cognitive Systems 2022

☆ Word Embedding Neural Networks to Advance Knee Osteoarthritis Research

Soheyla Amirian, Husam Ghazaleh, Mehdi Assefi, Hilal Maradit Kremers, Hamid R. Arabnia, Johannes F. Plate, Ahmad P. Tafti

Osteoarthritis (OA) is the most prevalent chronic joint disease worldwide, where knee OA takes more than 80% of commonly affected joints. Knee OA is not a curable disease yet, and it affects large columns of patients, making it costly to patients and healthcare systems. Etiology, diagnosis, and treatment of knee OA might be argued by variability in its clinical and physical manifestations. Although knee OA carries a list of well-known terminology aiming to standardize the nomenclature of the diagnosis, prognosis, treatment, and clinical outcomes of the chronic joint disease, in practice there is a wide range of terminology associated with knee OA across different data sources, including but not limited to biomedical literature, clinical notes, healthcare literacy, and health-related social media. Among these data sources, the scientific articles published in the biomedical literature usually make a principled pipeline to study disease. Rapid yet, accurate text mining on large-scale scientific literature may discover novel knowledge and terminology to better understand knee OA and to improve the quality of knee OA diagnosis, prevention, and treatment. The present works aim to utilize artificial neural network strategies to automatically extract vocabularies associated with knee OA diseases. Our finding indicates the feasibility of developing word embedding neural networks for autonomous keyword extraction and abstraction of knee OA.

comment: 5 pages, 3 figures, Accepted in Computational Science and Computational Intelligence; 2022 International Conference on IEEE CPS (IEEE XPLORE, Scopus)

☆ Realizing Molecular Machine Learning through Communications for Biological AI: Future Directions and Challenges

Sasitharan Balasubramaniam, Samitha Somathilaka, Sehee Sun, Adrian Ratwatte, Massimiliano Pierobon

Artificial Intelligence (AI) and Machine Learning (ML) are weaving their way into the fabric of society, where they are playing a crucial role in numerous facets of our lives. As we witness the increased deployment of AI and ML in various types of devices, we benefit from their use into energy-efficient algorithms for low powered devices. In this paper, we investigate a scale and medium that is far smaller than conventional devices as we move towards molecular systems that can be utilized to perform machine learning functions, i.e., Molecular Machine Learning (MML). Fundamental to the operation of MML is the transport, processing, and interpretation of information propagated by molecules through chemical reactions. We begin by reviewing the current approaches that have been developed for MML, before we move towards potential new directions that rely on gene regulatory networks inside biological organisms as well as their population interactions to create neural networks. We then investigate mechanisms for training machine learning structures in biological cells based on calcium signaling and demonstrate their application to build an Analog to Digital Converter (ADC). Lastly, we look at potential future directions as well as challenges that this area could solve.

☆ Decoding surface codes with deep reinforcement learning and probabilistic policy reuse

Elisha Siddiqui Matekole, Esther Ye, Ramya Iyer, Samuel Yen-Chi Chen

Quantum computing (QC) promises significant advantages on certain hard computational tasks over classical computers. However, current quantum hardware, also known as noisy intermediate-scale quantum computers (NISQ), are still unable to carry out computations faithfully mainly because of the lack of quantum error correction (QEC) capability. A significant amount of theoretical studies have provided various types of QEC codes; one of the notable topological codes is the surface code, and its features, such as the requirement of only nearest-neighboring two-qubit control gates and a large error threshold, make it a leading candidate for scalable quantum computation. Recent developments of machine learning (ML)-based techniques especially the reinforcement learning (RL) methods have been applied to the decoding problem and have already made certain progress. Nevertheless, the device noise pattern may change over time, making trained decoder models ineffective. In this paper, we propose a continual reinforcement learning method to address these decoding challenges. Specifically, we implement double deep Q-learning with probabilistic policy reuse (DDQN-PPR) model to learn surface code decoding strategies for quantum environments with varying noise patterns. Through numerical simulations, we show that the proposed DDQN-PPR model can significantly reduce the computational complexity. Moreover, increasing the number of trained policies can further improve the agent's performance. Our results open a way to build more capable RL agents which can leverage previously gained knowledge to tackle QEC challenges.

☆ Impossibility Theorems for Feature Attribution

Blair Bilodeau, Natasha Jaques, Pang Wei Koh, Been Kim

Despite a sea of interpretability methods that can produce plausible explanations, the field has also empirically seen many failure cases of such methods. In light of these results, it remains unclear for practitioners how to use these methods and choose between them in a principled way. In this paper, we show that for even moderately rich model classes (easily satisfied by neural networks), any feature attribution method that is complete and linear--for example, Integrated Gradients and SHAP--can provably fail to improve on random guessing for inferring model behaviour. Our results apply to common end-tasks such as identifying local model behaviour, spurious feature identification, and algorithmic recourse. One takeaway from our work is the importance of concretely defining end-tasks. In particular, we show that once such an end-task is defined, a simple and direct approach of repeated model evaluations can outperform many other complex feature attribution methods.

comment: 33 pages, 4 figures

☆ StoRM: A Diffusion-based Stochastic Regeneration Model for Speech Enhancement and Dereverberation

Jean-Marie Lemercier, Julius Richter, Simon Welker, Timo Gerkmann

Diffusion models have shown a great ability at bridging the performance gap between predictive and generative approaches for speech enhancement. We have shown that they may even outperform their predictive counterparts for non-additive corruption types or when they are evaluated on mismatched conditions. However, diffusion models suffer from a high computational burden, mainly as they require to run a neural network for each reverse diffusion step, whereas predictive approaches only require one pass. As diffusion models are generative approaches they may also produce vocalizing and breathing artifacts in adverse conditions. In comparison, in such difficult scenarios, predictive models typically do not produce such artifacts but tend to distort the target speech instead, thereby degrading the speech quality. In this work, we present a stochastic regeneration approach where an estimate given by a predictive model is provided as a guide for further diffusion. We show that the proposed approach uses the predictive model to remove the vocalizing and breathing artifacts while producing very high quality samples thanks to the diffusion model, even in adverse conditions. We further show that this approach enables to use lighter sampling schemes with fewer diffusion steps without sacrificing quality, thus lifting the computational burden by an order of magnitude. Source code and audio examples are available online (https://uhh.de/inf-sp-storm).

comment: This work has been submitted to the IEEE for publication. Copyright may be transferred without notice

☆ Fully 3D Implementation of the End-to-end Deep Image Prior-based PET Image Reconstruction Using Block Iterative Algorithm

Fumio Hashimoto, Yuya Onishi, Kibo Ote, Hideaki Tashima, Taiga Yamaya

Deep image prior (DIP) has recently attracted attention owing to its unsupervised positron emission tomography (PET) image reconstruction, which does not require any prior training dataset. In this paper, we present the first attempt to implement an end-to-end DIP-based fully 3D PET image reconstruction method that incorporates a forward-projection model into a loss function. To implement a practical fully 3D PET image reconstruction, which could not be performed due to a graphics processing unit memory limitation, we modify the DIP optimization to block-iteration and sequentially learn an ordered sequence of block sinograms. Furthermore, the relative difference penalty (RDP) term was added to the loss function to enhance the quantitative PET image accuracy. We evaluated our proposed method using Monte Carlo simulation with [$^{18}$F]FDG PET data of a human brain and a preclinical study on monkey brain [$^{18}$F]FDG PET data. The proposed method was compared with the maximum-likelihood expectation maximization (EM), maximum-a-posterior EM with RDP, and hybrid DIP-based PET reconstruction methods. The simulation results showed that the proposed method improved the PET image quality by reducing statistical noise and preserved a contrast of brain structures and inserted tumor compared with other algorithms. In the preclinical experiment, finer structures and better contrast recovery were obtained by the proposed method. This indicated that the proposed method can produce high-quality images without a prior training dataset. Thus, the proposed method is a key enabling technology for the straightforward and practical implementation of end-to-end DIP-based fully 3D PET image reconstruction.

comment: 9 pages, 10 figures

☆ The Quantum Path Kernel: a Generalized Quantum Neural Tangent Kernel for Deep Quantum Machine Learning

Massimiliano Incudini, Michele Grossi, Antonio Mandarino, Sofia Vallecorsa, Alessandra Di Pierro, David Windridge

Building a quantum analog of classical deep neural networks represents a fundamental challenge in quantum computing. A key issue is how to address the inherent non-linearity of classical deep learning, a problem in the quantum domain due to the fact that the composition of an arbitrary number of quantum gates, consisting of a series of sequential unitary transformations, is intrinsically linear. This problem has been variously approached in the literature, principally via the introduction of measurements between layers of unitary transformations. In this paper, we introduce the Quantum Path Kernel, a formulation of quantum machine learning capable of replicating those aspects of deep machine learning typically associated with superior generalization performance in the classical domain, specifically, hierarchical feature learning. Our approach generalizes the notion of Quantum Neural Tangent Kernel, which has been used to study the dynamics of classical and quantum machine learning models. The Quantum Path Kernel exploits the parameter trajectory, i.e. the curve delineated by model parameters as they evolve during training, enabling the representation of differential layer-wise convergence behaviors, or the formation of hierarchical parametric dependencies, in terms of their manifestation in the gradient space of the predictor function. We evaluate our approach with respect to variants of the classification of Gaussian XOR mixtures - an artificial but emblematic problem that intrinsically requires multilevel learning in order to achieve optimal class separation.

☆ Renormalization in the neural network-quantum field theory correspondence NeurIPS 2022

Harold Erbin, Vincent Lahoche, Dine Ousmane Samary

A statistical ensemble of neural networks can be described in terms of a quantum field theory (NN-QFT correspondence). The infinite-width limit is mapped to a free field theory, while finite N corrections are mapped to interactions. After reviewing the correspondence, we will describe how to implement renormalization in this context and discuss preliminary numerical results for translation-invariant kernels. A major outcome is that changing the standard deviation of the neural network weight distribution corresponds to a renormalization flow in the space of networks.

comment: A shorter version of this paper has been accepted in the NeurIPS 2022 workshop: Machine learning and the physical sciences (https://ml4physicalsciences.github.io/2022/)

☆ GAN-based Domain Inference Attack AAAI23

Yuechun Gu, Keke Chen

Model-based attacks can infer training data information from deep neural network models. These attacks heavily depend on the attacker's knowledge of the application domain, e.g., using it to determine the auxiliary data for model-inversion attacks. However, attackers may not know what the model is used for in practice. We propose a generative adversarial network (GAN) based method to explore likely or similar domains of a target model -- the model domain inference (MDI) attack. For a given target (classification) model, we assume that the attacker knows nothing but the input and output formats and can use the model to derive the prediction for any input in the desired form. Our basic idea is to use the target model to affect a GAN training process for a candidate domain's dataset that is easy to obtain. We find that the target model may distract the training procedure less if the domain is more similar to the target domain. We then measure the distraction level with the distance between GAN-generated datasets, which can be used to rank candidate domains for the target model. Our experiments show that the auxiliary dataset from an MDI top-ranked domain can effectively boost the result of model-inversion attacks.

comment: accepted by AAAI23

☆ Device Selection for the Coexistence of URLLC and Distributed Learning Services

Milad Ganjalizadeh, Hossein Shokri Ghadikolaei, Deniz Gündüz, Marina Petrova

Recent advances in distributed artificial intelligence (AI) have led to tremendous breakthroughs in various communication services, from fault-tolerant factory automation to smart cities. When distributed learning is run over a set of wirelessly connected devices, random channel fluctuations and the incumbent services running on the same network impact the performance of both distributed learning and the coexisting service. In this paper, we investigate a mixed service scenario where distributed AI workflow and ultra-reliable low latency communication (URLLC) services run concurrently over a network. Consequently, we propose a risk sensitivity-based formulation for device selection to minimize the AI training delays during its convergence period while ensuring that the operational requirements of the URLLC service are met. To address this challenging coexistence problem, we transform it into a deep reinforcement learning problem and address it via a framework based on soft actor-critic algorithm. We evaluate our solution with a realistic and 3GPP-compliant simulator for factory automation use cases. Our simulation results confirm that our solution can significantly decrease the training delay of the distributed AI service while keeping the URLLC availability above its required threshold and close to the scenario where URLLC solely consumes all network resources.

comment: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

☆ EuclidNets: An Alternative Operation for Efficient Inference of Deep Learning Models

Xinlin Li, Mariana Parazeres, Adam Oberman, Alireza Ghaffari, Masoud Asgharian, Vahid Partovi Nia

With the advent of deep learning application on edge devices, researchers actively try to optimize their deployments on low-power and restricted memory devices. There are established compression method such as quantization, pruning, and architecture search that leverage commodity hardware. Apart from conventional compression algorithms, one may redesign the operations of deep learning models that lead to more efficient implementation. To this end, we propose EuclidNet, a compression method, designed to be implemented on hardware which replaces multiplication, $xw$, with Euclidean distance $(x-w)^2$. We show that EuclidNet is aligned with matrix multiplication and it can be used as a measure of similarity in case of convolutional layers. Furthermore, we show that under various transformations and noise scenarios, EuclidNet exhibits the same performance compared to the deep learning models designed with multiplication operations.

☆ Training Integer-Only Deep Recurrent Neural Networks

Vahid Partovi Nia, Eyyüb Sari, Vanessa Courville, Masoud Asgharian

Recurrent neural networks (RNN) are the backbone of many text and speech applications. These architectures are typically made up of several computationally complex components such as; non-linear activation functions, normalization, bi-directional dependence and attention. In order to maintain good accuracy, these components are frequently run using full-precision floating-point computation, making them slow, inefficient and difficult to deploy on edge devices. In addition, the complex nature of these operations makes them challenging to quantize using standard quantization methods without a significant performance drop. We present a quantization-aware training method for obtaining a highly accurate integer-only recurrent neural network (iRNN). Our approach supports layer normalization, attention, and an adaptive piecewise linear (PWL) approximation of activation functions, to serve a wide range of state-of-the-art RNNs. The proposed method enables RNN-based language models to run on edge devices with $2\times$ improvement in runtime, and $4\times$ reduction in model size while maintaining similar accuracy as its full-precision counterpart.

comment: arXiv admin note: substantial text overlap with arXiv:2109.09828

☆ Adversarial Machine Learning and Defense Game for NextG Signal Classification with Deep Learning

Yalin E. Sagduyu

This paper presents a game-theoretic framework to study the interactions of attack and defense for deep learning-based NextG signal classification. NextG systems such as the one envisioned for a massive number of IoT devices can employ deep neural networks (DNNs) for various tasks such as user equipment identification, physical layer authentication, and detection of incumbent users (such as in the Citizens Broadband Radio Service (CBRS) band). By training another DNN as the surrogate model, an adversary can launch an inference (exploratory) attack to learn the behavior of the victim model, predict successful operation modes (e.g., channel access), and jam them. A defense mechanism can increase the adversary's uncertainty by introducing controlled errors in the victim model's decisions (i.e., poisoning the adversary's training data). This defense is effective against an attack but reduces the performance when there is no attack. The interactions between the defender and the adversary are formulated as a non-cooperative game, where the defender selects the probability of defending or the defense level itself (i.e., the ratio of falsified decisions) and the adversary selects the probability of attacking. The defender's objective is to maximize its reward (e.g., throughput or transmission success ratio), whereas the adversary's objective is to minimize this reward and its attack cost. The Nash equilibrium strategies are determined as operation modes such that no player can unilaterally improve its utility given the other's strategy is fixed. A fictitious play is formulated for each player to play the game repeatedly in response to the empirical frequency of the opponent's actions. The performance in Nash equilibrium is compared to the fixed attack and defense cases, and the resilience of NextG signal classification against attacks is quantified.

☆ Fixed-budget online adaptive mesh learning for physics-informed neural networks. Towards parameterized problem inference

Thi Nguyen Khoa Nguyen, Thibault Dairay, Raphaël Meunier, Christophe Millet, Mathilde Mougeot

Physics-Informed Neural Networks (PINNs) have gained much attention in various fields of engineering thanks to their capability of incorporating physical laws into the models. PINNs integrate the physical constraints by minimizing the partial differential equations (PDEs) residuals on a set of collocation points. The distribution of these collocation points appears to have a huge impact on the performance of PINNs and the assessment of the sampling methods for these points is still an active topic. In this paper, we propose a Fixed-Budget Online Adaptive Mesh Learning (FBOAML) method, which decomposes the domain into sub-domains, for training collocation points based on local maxima and local minima of the PDEs residuals. The stopping criterion is based on a data set of reference, which leads to an adaptive number of iterations for each specific problem. The effectiveness of FBOAML is demonstrated in the context of non-parameterized and parameterized problems. The impact of the hyper-parameters in FBOAML is investigated in this work. The comparison with other adaptive sampling methods is also illustrated. The numerical results demonstrate important gains in terms of accuracy of PINNs with FBOAML over the classical PINNs with non-adaptive collocation points. We also apply FBOAML in a complex industrial application involving coupling between mechanical and thermal fields. We show that FBOAML is able to identify the high-gradient location and even give better prediction for some physical fields than the classical PINNs with collocation points taken on a pre-adapted finite element mesh.

comment: 15 pages, 16 figures, 2 tables

☆ Few-shot human motion prediction for heterogeneous sensors

Rafael Rego Drumond, Lukas Brinkmeyer, Lars Schmidt-Thieme

Human motion prediction is a complex task as it involves forecasting variables over time on a graph of connected sensors. This is especially true in the case of few-shot learning, where we strive to forecast motion sequences for previously unseen actions based on only a few examples. Despite this, almost all related approaches for few-shot motion prediction do not incorporate the underlying graph, while it is a common component in classical motion prediction. Furthermore, state-of-the-art methods for few-shot motion prediction are restricted to motion tasks with a fixed output space meaning these tasks are all limited to the same sensor graph. In this work, we propose to extend recent works on few-shot time-series forecasting with heterogeneous attributes with graph neural networks to introduce the first few-shot motion approach that explicitly incorporates the spatial graph while also generalizing across motion tasks with heterogeneous sensors. In our experiments on motion tasks with heterogeneous sensors, we demonstrate significant performance improvements with lifts from 10.4% up to 39.3% compared to best state-of-the-art models. Moreover, we show that our model can perform on par with the best approach so far when evaluating on tasks with a fixed output space while maintaining two magnitudes fewer parameters.

☆ Mind Your Heart: Stealthy Backdoor Attack on Dynamic Deep Neural Network in Edge Computing

Tian Dong, Ziyuan Zhang, Han Qiu, Tianwei Zhang, Hewu Li, Terry Wang

Transforming off-the-shelf deep neural network (DNN) models into dynamic multi-exit architectures can achieve inference and transmission efficiency by fragmenting and distributing a large DNN model in edge computing scenarios (e.g., edge devices and cloud servers). In this paper, we propose a novel backdoor attack specifically on the dynamic multi-exit DNN models. Particularly, we inject a backdoor by poisoning one DNN model's shallow hidden layers targeting not this vanilla DNN model but only its dynamically deployed multi-exit architectures. Our backdoored vanilla model behaves normally on performance and cannot be activated even with the correct trigger. However, the backdoor will be activated when the victims acquire this model and transform it into a dynamic multi-exit architecture at their deployment. We conduct extensive experiments to prove the effectiveness of our attack on three structures (ResNet-56, VGG-16, and MobileNet) with four datasets (CIFAR-10, SVHN, GTSRB, and Tiny-ImageNet) and our backdoor is stealthy to evade multiple state-of-the-art backdoor detection or removal methods.

comment: Accepted to IEEE INFOCOM 2023

☆ Certified Policy Smoothing for Cooperative Multi-Agent Reinforcement Learning AAAI2023

Ronghui Mu, Wenjie Ruan, Leandro Soriano Marcolino, Gaojie Jin, Qiang Ni

Cooperative multi-agent reinforcement learning (c-MARL) is widely applied in safety-critical scenarios, thus the analysis of robustness for c-MARL models is profoundly important. However, robustness certification for c-MARLs has not yet been explored in the community. In this paper, we propose a novel certification method, which is the first work to leverage a scalable approach for c-MARLs to determine actions with guaranteed certified bounds. c-MARL certification poses two key challenges compared with single-agent systems: (i) the accumulated uncertainty as the number of agents increases; (ii) the potential lack of impact when changing the action of a single agent into a global team reward. These challenges prevent us from directly using existing algorithms. Hence, we employ the false discovery rate (FDR) controlling procedure considering the importance of each agent to certify per-state robustness and propose a tree-search-based algorithm to find a lower bound of the global reward under the minimal certified perturbation. As our method is general, it can also be applied in single-agent environments. We empirically show that our certification bounds are much tighter than state-of-the-art RL certification solutions. We also run experiments on two popular c-MARL algorithms: QMIX and VDN, in two different environments, with two and four agents. The experimental results show that our method produces meaningful guaranteed robustness for all models and environments. Our tool CertifyCMARL is available at https://github.com/TrustAI/CertifyCMA

comment: This paper will appear in AAAI2023

☆ The State of the Art in Enhancing Trust in Machine Learning Models with the Use of Visualizations

A. Chatzimparmpas, R. Martins, I. Jusufi, K. Kucher, Fabrice Rossi, A. Kerren

Machine learning (ML) models are nowadays used in complex applications in various domains, such as medicine, bioinformatics, and other sciences. Due to their black box nature, however, it may sometimes be hard to understand and trust the results they provide. This has increased the demand for reliable visualization tools related to enhancing trust in ML models, which has become a prominent topic of research in the visualization community over the past decades. To provide an overview and present the frontiers of current research on the topic, we present a State-of-the-Art Report (STAR) on enhancing trust in ML models with the use of interactive visualization. We define and describe the background of the topic, introduce a categorization for visualization techniques that aim to accomplish this goal, and discuss insights and opportunities for future research directions. Among our contributions is a categorization of trust against different facets of interactive ML, expanded and improved from previous research. Our results are investigated from different analytical perspectives: (a) providing a statistical overview, (b) summarizing key findings, (c) performing topic analyses, and (d) exploring the data sets used in the individual papers, all with the support of an interactive web-based survey browser. We intend this survey to be beneficial for visualization researchers whose interests involve making ML models more trustworthy, as well as researchers and practitioners from other disciplines in their search for effective visualization techniques suitable for solving their tasks with confidence and conveying meaning to their data.

☆ TransPath: Learning Heuristics For Grid-Based Pathfinding via Transformers AAAI'23

Daniil Kirilenko, Anton Andreychuk, Aleksandr Panov, Konstantin Yakovlev

Heuristic search algorithms, e.g. A*, are the commonly used tools for pathfinding on grids, i.e. graphs of regular structure that are widely employed to represent environments in robotics, video games etc. Instance-independent heuristics for grid graphs, e.g. Manhattan distance, do not take the obstacles into account and, thus, the search led by such heuristics performs poorly in the obstacle-rich environments. To this end, we suggest learning the instance-dependent heuristic proxies that are supposed to notably increase the efficiency of the search. The first heuristic proxy we suggest to learn is the correction factor, i.e. the ratio between the instance independent cost-to-go estimate and the perfect one (computed offline at the training phase). Unlike learning the absolute values of the cost-to-go heuristic function, which was known before, when learning the correction factor the knowledge of the instance-independent heuristic is utilized. The second heuristic proxy is the path probability, which indicates how likely the grid cell is lying on the shortest path. This heuristic can be utilized in the Focal Search framework as the secondary heuristic, allowing us to preserve the guarantees on the bounded sub-optimality of the solution. We learn both suggested heuristics in a supervised fashion with the state-of-the-art neural networks containing attention blocks (transformers). We conduct a thorough empirical evaluation on a comprehensive dataset of planning tasks, showing that the suggested techniques i) reduce the computational effort of the A* up to a factor of $4$x while producing the solutions, which costs exceed the costs of the optimal solutions by less than $0.3$% on average; ii) outperform the competitors, which include the conventional techniques from the heuristic search, i.e. weighted A*, as well as the state-of-the-art learnable planners.

comment: Pre-print of the paper accepted to AAAI'23

☆ Federated Learning -- Methods, Applications and beyond

Moritz Heusinger, Christoph Raab, Fabrice Rossi, Frank-Michael Schleif

In recent years the applications of machine learning models have increased rapidly, due to the large amount of available data and technological progress.While some domains like web analysis can benefit from this with only minor restrictions, other fields like in medicine with patient data are strongerregulated. In particular \emph{data privacy} plays an important role as recently highlighted by the trustworthy AI initiative of the EU or general privacy regulations in legislation. Another major challenge is, that the required training \emph{data is} often \emph{distributed} in terms of features or samples and unavailable for classicalbatch learning approaches. In 2016 Google came up with a framework, called \emph{Federated Learning} to solve both of these problems. We provide a brief overview on existing Methods and Applications in the field of vertical and horizontal \emph{Federated Learning}, as well as \emph{Fderated Transfer Learning}.

☆ Co-clustering based exploratory analysis of mixed-type data tables

Aichetou Bouchareb, Marc Boullé, Fabrice Clérot, Fabrice Rossi

Co-clustering is a class of unsupervised data analysis techniques that extract the existing underlying dependency structure between the instances and variables of a data table as homogeneous blocks. Most of those techniques are limited to variables of the same type. In this paper, we propose a mixed data co-clustering method based on a two-step methodology. In the first step, all the variables are binarized according to a number of bins chosen by the analyst, by equal frequency discretization in the numerical case, or keeping the most frequent values in the categorical case. The second step applies a co-clustering to the instances and the binary variables, leading to groups of instances and groups of variable parts. We apply this methodology on several data sets and compare with the results of a Multiple Correspondence Analysis applied to the same data.

☆ Reusable Options through Gradient-based Meta Learning

David Kuric, Herke van Hoof

Hierarchical methods in reinforcement learning have the potential to reduce the amount of decisions that the agent needs to perform when learning new tasks. However, finding a reusable useful temporal abstractions that facilitate fast learning remains a challenging problem. Recently, several deep learning approaches were proposed to learn such temporal abstractions in the form of options in an end-to-end manner. In this work, we point out several shortcomings of these methods and discuss their potential negative consequences. Subsequently, we formulate the desiderata for reusable options and use these to frame the problem of learning options as a gradient-based meta-learning problem. This allows us to formulate an objective that explicitly incentivizes options which allow a higher-level decision maker to adjust in few steps to different tasks. Experimentally, we show that our method is able to learn transferable components which accelerate learning and performs better than existing prior methods developed for this setting. Additionally, we perform ablations to quantify the impact of using gradient-based meta-learning as well as other proposed changes.

☆ Model Based Co-clustering of Mixed Numerical and Binary Data

Aichetou Bouchareb, Marc Boullé, Fabrice Clérot, Fabrice Rossi

Co-clustering is a data mining technique used to extract the underlying block structure between the rows and columns of a data matrix. Many approaches have been studied and have shown their capacity to extract such structures in continuous, binary or contingency tables. However, very little work has been done to perform co-clustering on mixed type data. In this article, we extend the latent block models based co-clustering to the case of mixed data (continuous and binary variables). We then evaluate the effectiveness of the proposed approach on simulated data and we discuss its advantages and potential limits.

☆ GOOD: Exploring Geometric Cues for Detecting Objects in an Open World ICLR 2023

Haiwen Huang, Andreas Geiger, Dan Zhang

We address the task of open-world class-agnostic object detection, i.e., detecting every object in an image by learning from a limited number of base object classes. State-of-the-art RGB-based models suffer from overfitting the training classes and often fail at detecting novel-looking objects. This is because RGB-based models primarily rely on appearance similarity to detect novel objects and are also prone to overfitting short-cut cues such as textures and discriminative parts. To address these shortcomings of RGB-based object detectors, we propose incorporating geometric cues such as depth and normals, predicted by general-purpose monocular estimators. Specifically, we use the geometric cues to train an object proposal network for pseudo-labeling unannotated novel objects in the training set. Our resulting Geometry-guided Open-world Object Detector (GOOD) significantly improves detection recall for novel object categories and already performs well with only a few training classes. Using a single "person" class for training on the COCO dataset, GOOD surpasses SOTA methods by 5.0% AR@100, a relative improvement of 24%.

comment: Under review as a conference paper at ICLR 2023

☆ Reinforcement Learning Based Approaches to Adaptive Context Caching in Distributed Context Management Systems

Shakthi Weerasinghe, Arkady Zaslavsky, Seng W. Loke, Amin Abken, Alireza Hassani

Performance metrics-driven context caching has a profound impact on throughput and response time in distributed context management systems for real-time context queries. This paper proposes a reinforcement learning based approach to adaptively cache context with the objective of minimizing the cost incurred by context management systems in responding to context queries. Our novel algorithms enable context queries and sub-queries to reuse and repurpose cached context in an efficient manner. This approach is distinctive to traditional data caching approaches by three main features. First, we make selective context cache admissions using no prior knowledge of the context, or the context query load. Secondly, we develop and incorporate innovative heuristic models to calculate expected performance of caching an item when making the decisions. Thirdly, our strategy defines a time-aware continuous cache action space. We present two reinforcement learning agents, a value function estimating actor-critic agent and a policy search agent using deep deterministic policy gradient method. The paper also proposes adaptive policies such as eviction and cache memory scaling to complement our objective. Our method is evaluated using a synthetically generated load of context sub-queries and a synthetic data set inspired from real world data and query samples. We further investigate optimal adaptive caching configurations under different settings. This paper presents, compares, and discusses our findings that the proposed selective caching methods reach short- and long-term cost- and performance-efficiency. The paper demonstrates that the proposed methods outperform other modes of context management such as redirector mode, and database mode, and cache all policy by up to 60% in cost efficiency.

comment: This is a pre-print version of the journal paper submitted to ACM Transactions in Internet of Things, which is currently under review

☆ Robust Meta-Representation Learning via Global Label Inference and Classification

Ruohan Wang, Isak Falk, Massimiliano Pontil, Carlo Ciliberto

Few-shot learning (FSL) is a central problem in meta-learning, where learners must efficiently learn from few labeled examples. Within FSL, feature pre-training has recently become an increasingly popular strategy to significantly improve generalization performance. However, the contribution of pre-training is often overlooked and understudied, with limited theoretical understanding of its impact on meta-learning performance. Further, pre-training requires a consistent set of global labels shared across training tasks, which may be unavailable in practice. In this work, we address the above issues by first showing the connection between pre-training and meta-learning. We discuss why pre-training yields more robust meta-representation and connect the theoretical analysis to existing works and empirical results. Secondly, we introduce Meta Label Learning (MeLa), a novel meta-learning algorithm that learns task relations by inferring global labels across tasks. This allows us to exploit pre-training for FSL even when global labels are unavailable or ill-defined. Lastly, we introduce an augmented pre-training procedure that further improves the learned meta-representation. Empirically, MeLa outperforms existing methods across a diverse range of benchmarks, in particular under a more challenging setting where the number of training tasks is limited and labels are task-specific. We also provide extensive ablation study to highlight its key properties.

☆ GENIE: Large Scale Pre-training for Text Generation with Diffusion Model

Zhenghao Lin, Yeyun Gong, Yelong Shen, Tong Wu, Zhihao Fan, Chen Lin, Weizhu Chen, Nan Duan

In this paper, we propose a large-scale language pre-training for text GENeration using dIffusion modEl, which is named GENIE. GENIE is a pre-training sequence-to-sequence text generation model which combines Transformer and diffusion. The diffusion model accepts the latent information from the encoder, which is used to guide the denoising of the current time step. After multiple such denoise iterations, the diffusion model can restore the Gaussian noise to the diverse output text which is controlled by the input text. Moreover, such architecture design also allows us to adopt large scale pre-training on the GENIE. We propose a novel pre-training method named continuous paragraph denoise based on the characteristics of the diffusion model. Extensive experiments on the XSum, CNN/DailyMail, and Gigaword benchmarks shows that GENIE can achieves comparable performance with various strong baselines, especially after pre-training, the generation quality of GENIE is greatly improved. We have also conduct a lot of experiments on the generation diversity and parameter impact of GENIE. The code for GENIE will be made publicly available.

comment: Work in progress

☆ Set-Transformer BeamsNet for AUV Velocity Forecasting in Complete DVL Outage Scenarios

Nadav Cohen, Zeev Yampolsky, Itzik Klein

Autonomous underwater vehicles (AUVs) are regularly used for deep ocean applications. Commonly, the autonomous navigation task is carried out by a fusion between two sensors: the inertial navigation system and the Doppler velocity log (DVL). The DVL operates by transmitting four acoustic beams to the sea floor, and once reflected back, the AUV velocity vector can be estimated. However, in real-life scenarios, such as an uneven seabed, sea creatures blocking the DVL's view and, roll/pitch maneuvers, the acoustic beams' reflection is resulting in a scenario known as DVL outage. Consequently, a velocity update is not available to bind the inertial solution drift. To cope with such situations, in this paper, we leverage our BeamsNet framework and propose a Set-Transformer-based BeamsNet (ST-BeamsNet) that utilizes inertial data readings and previous DVL velocity measurements to regress the current AUV velocity in case of a complete DVL outage. The proposed approach was evaluated using data from experiments held in the Mediterranean Sea with the Snapir AUV and was compared to a moving average (MA) estimator. Our ST-BeamsNet estimated the AUV velocity vector with an 8.547% speed error, which is 26% better than the MA approach.

☆ Towards Causal Credit Assignment

Mátyás Schubert

Adequately assigning credit to actions for future outcomes based on their contributions is a long-standing open challenge in Reinforcement Learning. The assumptions of the most commonly used credit assignment method are disadvantageous in tasks where the effects of decisions are not immediately evident. Furthermore, this method can only evaluate actions that have been selected by the agent, making it highly inefficient. Still, no alternative methods have been widely adopted in the field. Hindsight Credit Assignment is a promising, but still unexplored candidate, which aims to solve the problems of both long-term and counterfactual credit assignment. In this thesis, we empirically investigate Hindsight Credit Assignment to identify its main benefits, and key points to improve. Then, we apply it to factored state representations, and in particular to state representations based on the causal structure of the environment. In this setting, we propose a variant of Hindsight Credit Assignment that effectively exploits a given causal structure. We show that our modification greatly decreases the workload of Hindsight Credit Assignment, making it more efficient and enabling it to outperform the baseline credit assignment method on various tasks. This opens the way to other methods based on given or learned causal structures.

☆ Hybrid Quantum-Classical Generative Adversarial Network for High Resolution Image Generation

Shu Lok Tsang, Maxwell T. West, Sarah M. Erfani, Muhammad Usman

Quantum machine learning (QML) has received increasing attention due to its potential to outperform classical machine learning methods in various problems. A subclass of QML methods is quantum generative adversarial networks (QGANs) which have been studied as a quantum counterpart of classical GANs widely used in image manipulation and generation tasks. The existing work on QGANs is still limited to small-scale proof-of-concept examples based on images with significant down-scaling. Here we integrate classical and quantum techniques to propose a new hybrid quantum-classical GAN framework. We demonstrate its superior learning capabilities by generating $28 \times 28$ pixels grey-scale images without dimensionality reduction or classical pre/post-processing on multiple classes of the standard MNIST and Fashion MNIST datasets, which achieves comparable results to classical frameworks with 3 orders of magnitude less trainable generator parameters. To gain further insight into the working of our hybrid approach, we systematically explore the impact of its parameter space by varying the number of qubits, the size of image patches, the number of layers in the generator, the shape of the patches and the choice of prior distribution. Our results show that increasing the quantum generator size generally improves the learning capability of the network. The developed framework provides a foundation for future design of QGANs with optimal parameter set tailored for complex image generation tasks.

☆ Sequential Decision Problems with Weak Feedback

Arun Verma

This thesis considers sequential decision problems, where the loss/reward incurred by selecting an action may not be inferred from observed feedback. A major part of this thesis focuses on the unsupervised sequential selection problem, where one can not infer the loss incurred for selecting an action from observed feedback. We also introduce a new setup named Censored Semi Bandits, where the loss incurred for selecting an action can be observed under certain conditions. Finally, we study the channel selection problem in the communication networks, where the reward for an action is only observed when no other player selects that action to play in the round. These problems find applications in many fields like healthcare, crowd-sourcing, security, adaptive resource allocation, among many others. This thesis aims to address the above-described sequential decision problems by exploiting specific structures these problems exhibit. We develop provably optimal algorithms for each of these setups with weak feedback and validate their empirical performance on different problem instances derived from synthetic and real datasets.

comment: Ph.D. Thesis

☆ Synopsis: Sequential Decision Problems with Weak Feedback

Arun Verma

This thesis considers sequential decision problems, where the loss/reward incurred by selecting an action may not be inferred from observed feedback. A major part of this thesis focuses on the unsupervised sequential selection problem, where one can not infer the loss incurred for selecting an action from observed feedback. We also introduce a new setup named Censored Semi Bandits, where the loss incurred for selecting an action can be observed under certain conditions. Finally, we study the channel selection problem in the communication networks, where the reward for an action is only observed when no other player selects that action to play in the round. These problems find applications in many fields like healthcare, crowd-sourcing, security, adaptive resource allocation, among many others. This thesis aims to address the above-described sequential decision problems by exploiting specific structures these problems exhibit. We develop provably optimal algorithms for each of these setups with weak feedback and validate their empirical performance on different problem instances derived from synthetic and real datasets.

comment: arXiv admin note: text overlap with arXiv:2010.12353, arXiv:2009.07554

☆ AsyncFLEO: Asynchronous Federated Learning for LEO Satellite Constellations with High-Altitude Platforms

Mohamed Elmahallawy, Tie Luo

Low Earth Orbit (LEO) constellations, each comprising a large number of satellites, have become a new source of big data "from the sky". Downloading such data to a ground station (GS) for big data analytics demands very high bandwidth and involves large propagation delays. Federated Learning (FL) offers a promising solution because it allows data to stay in-situ (never leaving satellites) and it only needs to transmit machine learning model parameters (trained on the satellites' data). However, the conventional, synchronous FL process can take several days to train a single FL model in the context of satellite communication (Satcom), due to a bottleneck caused by straggler satellites. In this paper, we propose an asynchronous FL framework for LEO constellations called AsyncFLEO to improve FL efficiency in Satcom. Not only does AsynFLEO address the bottleneck (idle waiting) in synchronous FL, but it also solves the issue of model staleness caused by straggler satellites. AsyncFLEO utilizes high-altitude platforms (HAPs) positioned "in the sky" as parameter servers, and consists of three technical components: (1) a ring-of-stars communication topology, (2) a model propagation algorithm, and (3) a model aggregation algorithm with satellite grouping and staleness discounting. Our extensive evaluation with both IID and non-IID data shows that AsyncFLEO outperforms the state of the art by a large margin, cutting down convergence delay by 22 times and increasing accuracy by 40%.

comment: 2022 IEEE International Conference on Big Data (IEEE BigData 2022), December 2022

☆ Supervised Anomaly Detection Method Combining Generative Adversarial Networks and Three-Dimensional Data in Vehicle Inspections

Yohei Baba, Takuro Hoshi, Ryosuke Mori, Gaurang Gavai

The external visual inspections of rolling stock's underfloor equipment are currently being performed via human visual inspection. In this study, we attempt to partly automate visual inspection by investigating anomaly inspection algorithms that use image processing technology. As the railroad maintenance studies tend to have little anomaly data, unsupervised learning methods are usually preferred for anomaly detection; however, training cost and accuracy is still a challenge. Additionally, a researcher created anomalous images from normal images by adding noise, etc., but the anomalous targeted in this study is the rotation of piping cocks that was difficult to create using noise. Therefore, in this study, we propose a new method that uses style conversion via generative adversarial networks on three-dimensional computer graphics and imitates anomaly images to apply anomaly detection based on supervised learning. The geometry-consistent style conversion model was used to convert the image, and because of this the color and texture of the image were successfully made to imitate the real image while maintaining the anomalous shape. Using the generated anomaly images as supervised data, the anomaly detection model can be easily trained without complex adjustments and successfully detects anomalies.

comment: 6 pages, 12 figures

☆ Accelerating Barnes-Hut t-SNE Algorithm by Efficient Parallelization on Multi-Core CPUs

Narendra Chaudhary, Alexander Pivovar, Pavel Yakovlev, Andrey Gorshkov, Sanchit Misra

t-SNE remains one of the most popular embedding techniques for visualizing high-dimensional data. Most standard packages of t-SNE, such as scikit-learn, use the Barnes-Hut t-SNE (BH t-SNE) algorithm for large datasets. However, existing CPU implementations of this algorithm are inefficient. In this work, we accelerate the BH t-SNE on CPUs via cache optimizations, SIMD, parallelizing sequential steps, and improving parallelization of multithreaded steps. Our implementation (Acc-t-SNE) is up to 261x and 4x faster than scikit-learn and the state-of-the-art BH t-SNE implementation from daal4py, respectively, on a 32-core Intel(R) Icelake cloud instance.

☆ Scalable Multi-Agent Reinforcement Learning for Warehouse Logistics with Robotic and Human Co-Workers

Aleksandar Krnjaic, Jonathan D. Thomas, Georgios Papoudakis, Lukas Schäfer, Peter Börsting, Stefano V. Albrecht

This project leverages advances in multi-agent reinforcement learning (MARL) to improve the efficiency and flexibility of order-picking systems for commercial warehouses. We envision a warehouse of the future in which dozens of mobile robots and human pickers work together to collect and deliver items within the warehouse. The fundamental problem we tackle, called the order-picking problem, is how these worker agents must coordinate their movement and actions in the warehouse to maximise performance (e.g. order throughput) under given resource constraints. Established industry methods using heuristic approaches require large engineering efforts to optimise for innately variable warehouse configurations. In contrast, the MARL framework can be flexibly applied to any warehouse configuration (e.g. size, layout, number/types of workers, item replenishment frequency) and the agents learn via a process of trial-and-error how to optimally cooperate with one another. This paper details the current status of the R&D effort initiated by Dematic and the University of Edinburgh towards a general-purpose and scalable MARL solution for the order-picking problem in realistic warehouses.

☆ Understanding and Improving the Role of Projection Head in Self-Supervised Learning

Kartik Gupta, Thalaiyasingam Ajanthan, Anton van den Hengel, Stephen Gould

Self-supervised learning (SSL) aims to produce useful feature representations without access to any human-labeled data annotations. Due to the success of recent SSL methods based on contrastive learning, such as SimCLR, this problem has gained popularity. Most current contrastive learning approaches append a parametrized projection head to the end of some backbone network to optimize the InfoNCE objective and then discard the learned projection head after training. This raises a fundamental question: Why is a learnable projection head required if we are to discard it after training? In this work, we first perform a systematic study on the behavior of SSL training focusing on the role of the projection head layers. By formulating the projection head as a parametric component for the InfoNCE objective rather than a part of the network, we present an alternative optimization scheme for training contrastive learning based SSL frameworks. Our experimental study on multiple image classification datasets demonstrates the effectiveness of the proposed approach over alternatives in the SSL literature.

☆ Learning to swim efficiently in a nonuniform flow field

Krongtum Sankaewtong, John J. Molina, Matthew S. Turner, Ryoichi Yamamoto

Microswimmers can acquire information on the surrounding fluid by sensing mechanical queues. They can then navigate in response to these signals. We analyse this navigation by combining deep reinforcement learning with direct numerical simulations to resolve the hydrodynamics. We study how local and non-local information can be used to train a swimmer to achieve particular swimming tasks in a non-uniform flow field, in particular a zig-zag shear flow. The swimming tasks are (1) learning how to swim in the vorticity direction, (2) the shear-gradient direction, and (3) the shear flow direction. We find that access to lab frame information on the swimmer's instantaneous orientation is all that is required in order to reach the optimal policy for (1,2). However, information on both the translational and rotational velocities seem to be required to achieve (3). Inspired by biological microorganisms we also consider the case where the swimmers sense local information, i.e. surface hydrodynamic forces, together with a signal direction. This might correspond to gravity or, for micro-organisms with light sensors, a light source. In this case, we show that the swimmer can reach a comparable level of performance as a swimmer with access to lab frame variables. We also analyse the role of different swimming modes, i.e. pusher, puller, and neutral swimmers.

☆ A Mathematical Framework for Learning Probability Distributions

Hongkang Yang

The modeling of probability distributions, specifically generative modeling and density estimation, has become an immensely popular subject in recent years by virtue of its outstanding performance on sophisticated data such as images and texts. Nevertheless, a theoretical understanding of its success is still incomplete. One mystery is the paradox between memorization and generalization: In theory, the model is trained to be exactly the same as the empirical distribution of the finite samples, whereas in practice, the trained model can generate new samples or estimate the likelihood of unseen samples. Likewise, the overwhelming diversity of distribution learning models calls for a unified perspective on this subject. This paper provides a mathematical framework such that all the well-known models can be derived based on simple principles. To demonstrate its efficacy, we present a survey of our results on the approximation error, training error and generalization error of these models, which can all be established based on this framework. In particular, the aforementioned paradox is resolved by proving that these models enjoy implicit regularization during training, so that the generalization error at early-stopping avoids the curse of dimensionality. Furthermore, we provide some new results on landscape analysis and the mode collapse phenomenon.

☆ A machine learning framework for neighbor generation in metaheuristic search

Defeng Liu, Vincent Perreault, Alain Hertz, Andrea Lodi

This paper presents a methodology for integrating machine learning techniques into metaheuristics for solving combinatorial optimization problems. Namely, we propose a general machine learning framework for neighbor generation in metaheuristic search. We first define an efficient neighborhood structure constructed by applying a transformation to a selected subset of variables from the current solution. Then, the key of the proposed methodology is to generate promising neighbors by selecting a proper subset of variables that contains a descent of the objective in the solution space. To learn a good variable selection strategy, we formulate the problem as a classification task that exploits structural information from the characteristics of the problem and from high-quality solutions. We validate our methodology on two metaheuristic applications: a Tabu Search scheme for solving a Wireless Network Optimization problem and a Large Neighborhood Search heuristic for solving Mixed-Integer Programs. The experimental results show that our approach is able to achieve a satisfactory trade-off between the exploration of a larger solution space and the exploitation of high-quality solution regions on both applications.

☆ Commitment with Signaling under Double-sided Information Asymmetry

Tao Li, Quanyan Zhu

Information asymmetry in games enables players with the information advantage to manipulate others' beliefs by strategically revealing information to other players. This work considers a double-sided information asymmetry in a Bayesian Stackelberg game, where the leader's realized action, sampled from the mixed strategy commitment, is hidden from the follower. In contrast, the follower holds private information about his payoff. Given asymmetric information on both sides, an important question arises: \emph{Does the leader's information advantage outweigh the follower's?} We answer this question affirmatively in this work, where we demonstrate that by adequately designing a signaling device that reveals partial information regarding the leader's realized action to the follower, the leader can achieve a higher expected utility than that without signaling. Moreover, unlike previous works on the Bayesian Stackelberg game where mathematical programming tools are utilized, we interpret the leader's commitment as a probability measure over the belief space. Such a probabilistic language greatly simplifies the analysis and allows an indirect signaling scheme, leading to a geometric characterization of the equilibrium under the proposed game model.

comment: Working paper; 18 pages

☆ Offline Clustering Approach to Self-supervised Learning for Class-imbalanced Image Data

Hye-min Chang, Sungkyun Chang

Class-imbalanced datasets are known to cause the problem of model being biased towards the majority classes. In this project, we set up two research questions: 1) when is the class-imbalance problem more prevalent in self-supervised pre-training? and 2) can offline clustering of feature representations help pre-training on class-imbalanced data? Our experiments investigate the former question by adjusting the degree of {\it class-imbalance} when training the baseline models, namely SimCLR and SimSiam on CIFAR-10 database. To answer the latter question, we train each expert model on each subset of the feature clusters. We then distill the knowledge of expert models into a single model, so that we will be able to compare the performance of this model to our baselines.

comment: 5 pages, 3 figures, Technical report

☆ Novel Deep Learning Framework For Bovine Iris Segmentation

Heemoon Yoon, Mira Park, Sang-Hee Lee

Iris segmentation is the initial step to identify biometric of animals to establish a traceability system of livestock. In this study, we propose a novel deep learning framework for pixel-wise segmentation with minimum use of annotation labels using BovineAAEyes80 public dataset. In the experiment, U-Net with VGG16 backbone was selected as the best combination of encoder and decoder model, demonstrating a 99.50% accuracy and a 98.35% Dice coefficient score. Remarkably, the selected model accurately segmented corrupted images even without proper annotation data. This study contributes to the advancement of the iris segmentation and the development of a reliable DNNs training framework.

comment: 5 pages, 4 figures, 3 tables

☆ Local Policy Improvement for Recommender Systems

Dawen Liang, Nikos Vlassis

Recommender systems aim to answer the following question: given the items that a user has interacted with, what items will this user likely interact with next? Historically this problem is often framed as a predictive task via (self-)supervised learning. In recent years, we have seen more emphasis placed on approaching the recommendation problem from a policy optimization perspective: learning a policy that maximizes some reward function (e.g., user engagement). However, it is commonly the case in recommender systems that we are only able to train a new policy given data collected from a previously-deployed policy. The conventional way to address such a policy mismatch is through importance sampling correction, which unfortunately comes with its own limitations. In this paper, we suggest an alternative approach, which involves the use of local policy improvement without off-policy correction. Drawing from a number of related results in the fields of causal inference, bandits, and reinforcement learning, we present a suite of methods that compute and optimize a lower bound of the expected reward of the target policy. Crucially, this lower bound is a function that is easy to estimate from data, and which does not involve density ratios (such as those appearing in importance sampling correction). We argue that this local policy improvement paradigm is particularly well suited for recommender systems, given that in practice the previously-deployed policy is typically of reasonably high quality, and furthermore it tends to be re-trained frequently and gets continuously updated. We discuss some practical recipes on how to apply some of the proposed techniques in a sequential recommendation setting.

☆ Automatically Bounding the Taylor Remainder Series: Tighter Bounds and New Applications

Matthew Streeter, Joshua V. Dillon

We present a new algorithm for automatically bounding the Taylor remainder series. In the special case of a scalar function $f: \mathbb{R} \mapsto \mathbb{R}$, our algorithm takes as input a reference point $x_0$, trust region $[a, b]$, and integer $k \ge 0$, and returns an interval $I$ such that $f(x) - \sum_{i=0}^k \frac {f^{(i)}(x_0)} {i!} (x - x_0)^i \in I (x - x_0)^{k+1}$ for all $x \in [a, b]$. As in automatic differentiation, the function $f$ is provided to the algorithm in symbolic form, and must be composed of known elementary functions. At a high level, our algorithm has two steps. First, for a variety of commonly-used elementary functions (e.g., $\exp$, $\log$), we derive sharp polynomial upper and lower bounds on the Taylor remainder series. We then recursively combine the bounds for the elementary functions using an interval arithmetic variant of Taylor-mode automatic differentiation. Our algorithm can make efficient use of machine learning hardware accelerators, and we provide an open source implementation in JAX. We then turn our attention to applications. Most notably, we use our new machinery to create the first universal majorization-minimization optimization algorithms: algorithms that iteratively minimize an arbitrary loss using a majorizer that is derived automatically, rather than by hand. Applied to machine learning, this leads to architecture-specific optimizers for training deep networks that converge from any starting point, without hyperparameter tuning. Our experiments show that for some optimization problems, these hyperparameter-free optimizers outperform tuned versions of gradient descent, Adam, and AdaGrad. We also show that our automatically-derived bounds can be used for verified global optimization and numerical integration, and to prove sharper versions of Jensen's inequality.

comment: 85 pages, 17 figures

♻ ☆ Automatically Answering and Generating Machine Learning Final Exams

Sarah Zhang, Reece Shuttleworth, Zad Chin, Pedro Lantigua, Saisamrit Surbehera, Gregory Hunter, Derek Austin, Yann Hicke, Leonard Tang, Sathwik Karnik, Darnell Granberry, Iddo Drori

Can a machine learn machine learning? We propose to answer this question using the same criteria we use to answer a similar question: can a human learn machine learning? We automatically answer final exams in MIT's, Harvard's and Cornell's large machine learning courses and generate new questions at a human level. Recently, program synthesis and few-shot learning solved university-level problem set questions in mathematics and STEM courses at a human level. In this work, we solve questions from final exams that differ from problem sets in several ways: the questions are longer, have multiple parts, are more complicated, and span a broader set of topics. We provide a new dataset and benchmark of questions from machine learning final exams and code for automatically answering these questions and generating new questions. To make our dataset a reproducible benchmark, we use automatic checkers for multiple choice questions, questions with numeric answers, and questions with expression answers, and evaluate a large free language model, Meta's OPT, and compare the results with Open AI's GPT-3, ChatGPT, and Codex. A student survey comparing the quality, appropriateness, and difficulty of machine-generated questions with human-written questions shows that across multiple aspects, machine-generated questions are indistinguishable from human-generated questions and are suitable for final exams. We perform ablation studies comparing zero-shot learning with few-shot learning, chain-of-thought prompting, GPT-3, ChatGPT, and OPT pre-trained on text and Codex fine-tuned on code on a range of machine learning topics and find that few-shot learning methods perform best. We make our data and code publicly available for the machine learning community.

comment: 17 pages

♻ ☆ MetaFormer Baselines for Vision

Weihao Yu, Chenyang Si, Pan Zhou, Mi Luo, Yichen Zhou, Jiashi Feng, Shuicheng Yan, Xinchao Wang

MetaFormer, the abstracted architecture of Transformer, has been found to play a significant role in achieving competitive performance. In this paper, we further explore the capacity of MetaFormer, again, without focusing on token mixer design: we introduce several baseline models under MetaFormer using the most basic or common mixers, and summarize our observations as follows: (1) MetaFormer ensures solid lower bound of performance. By merely adopting identity mapping as the token mixer, the MetaFormer model, termed IdentityFormer, achieves >80% accuracy on ImageNet-1K. (2) MetaFormer works well with arbitrary token mixers. When specifying the token mixer as even a random matrix to mix tokens, the resulting model RandFormer yields an accuracy of >81%, outperforming IdentityFormer. Rest assured of MetaFormer's results when new token mixers are adopted. (3) MetaFormer effortlessly offers state-of-the-art results. With just conventional token mixers dated back five years ago, the models instantiated from MetaFormer already beat state of the art. (a) ConvFormer outperforms ConvNeXt. Taking the common depthwise separable convolutions as the token mixer, the model termed ConvFormer, which can be regarded as pure CNNs, outperforms the strong CNN model ConvNeXt. (b) CAFormer sets new record on ImageNet-1K. By simply applying depthwise separable convolutions as token mixer in the bottom stages and vanilla self-attention in the top stages, the resulting model CAFormer sets a new record on ImageNet-1K: it achieves an accuracy of 85.5% at 224x224 resolution, under normal supervised training without external data or distillation. In our expedition to probe MetaFormer, we also find that a new activation, StarReLU, reduces 71% FLOPs of activation compared with GELU yet achieves better performance. We expect StarReLU to find great potential in MetaFormer-like models alongside other neural networks.

comment: Add more ImageNet-22K pretrained models. Code: https://github.com/sail-sg/metaformer

♻ ☆ Learned Systems Security

Roei Schuster, Jin Peng Zhou, Thorsten Eisenhofer, Paul Grubbs, Nicolas Papernot

A learned system uses machine learning (ML) internally to improve performance. We can expect such systems to be vulnerable to some adversarial-ML attacks. Often, the learned component is shared between mutually-distrusting users or processes, much like microarchitectural resources such as caches, potentially giving rise to highly-realistic attacker models. However, compared to attacks on other ML-based systems, attackers face a level of indirection as they cannot interact directly with the learned model. Additionally, the difference between the attack surface of learned and non-learned versions of the same system is often subtle. These factors obfuscate the de-facto risks that the incorporation of ML carries. We analyze the root causes of potentially-increased attack surface in learned systems and develop a framework for identifying vulnerabilities that stem from the use of ML. We apply our framework to a broad set of learned systems under active development. To empirically validate the many vulnerabilities surfaced by our framework, we choose 3 of them and implement and evaluate exploits against prominent learned-system instances. We show that the use of ML caused leakage of past queries in a database, enabled a poisoning attack that causes exponential memory blowup in an index structure and crashes it in seconds, and enabled index users to snoop on each others' key distributions by timing queries over their own keys. We find that adversarial ML is a universal threat against learned systems, point to open research gaps in our understanding of learned-systems security, and conclude by discussing mitigations, while noting that data leakage is inherent in systems whose learned component is shared between multiple parties.

♻ ☆ Chatbots in a Botnet World

Forrest McKee, David Noever

Question-and-answer formats provide a novel experimental platform for investigating cybersecurity questions. Unlike previous chatbots, the latest ChatGPT model from OpenAI supports an advanced understanding of complex coding questions. The research demonstrates thirteen coding tasks that generally qualify as stages in the MITRE ATT&CK framework, ranging from credential access to defense evasion. With varying success, the experimental prompts generate examples of keyloggers, logic bombs, obfuscated worms, and payment-fulfilled ransomware. The empirical results illustrate cases that support the broad gain of functionality, including self-replication and self-modification, evasion, and strategic understanding of complex cybersecurity goals. One surprising feature of ChatGPT as a language-only model centers on its ability to spawn coding approaches that yield images that obfuscate or embed executable programming steps or links.

♻ ☆ EDICT: Exact Diffusion Inversion via Coupled Transformations

Bram Wallace, Akash Gokul, Nikhil Naik

Finding an initial noise vector that produces an input image when fed into the diffusion process (known as inversion) is an important problem in denoising diffusion models (DDMs), with applications for real image editing. The state-of-the-art approach for real image editing with inversion uses denoising diffusion implicit models (DDIMs) to deterministically noise the image to the intermediate state along the path that the denoising would follow given the original conditioning. However, DDIM inversion for real images is unstable as it relies on local linearization assumptions, which result in the propagation of errors, leading to incorrect image reconstruction and loss of content. To alleviate these problems, we propose Exact Diffusion Inversion via Coupled Transformations (EDICT), an inversion method that draws inspiration from affine coupling layers. EDICT enables mathematically exact inversion of real and model-generated images by maintaining two coupled noise vectors which are used to invert each other in an alternating fashion. Using Stable Diffusion, a state-of-the-art latent diffusion model, we demonstrate that EDICT successfully reconstructs real images with high fidelity. On complex image datasets like MS-COCO, EDICT reconstruction significantly outperforms DDIM, improving the mean square error of reconstruction by a factor of two. Using noise vectors inverted from real images, EDICT enables a wide range of image edits--from local and global semantic edits to image stylization--while maintaining fidelity to the original image structure. EDICT requires no model training/finetuning, prompt tuning, or extra data and can be combined with any pretrained DDM. Code is available at https://github.com/salesforce/EDICT.

comment: 24 pages, 22 figures. Code now available

♻ ☆ Towards mapping the contemporary art world with ArtLM: an art-specific NLP model

Qinkai Chen, Mohamed El-Mennaoui, Antoine Fosset, Amine Rebei, Haoyang Cao, Philine Bouscasse, Christy Eóin O'Beirne, Sasha Shevchenko, Mathieu Rosenbaum

With an increasing amount of data in the art world, discovering artists and artworks suitable to collectors' tastes becomes a challenge. It is no longer enough to use visual information, as contextual information about the artist has become just as important in contemporary art. In this work, we present a generic Natural Language Processing framework (called ArtLM) to discover the connections among contemporary artists based on their biographies. In this approach, we first continue to pre-train the existing general English language models with a large amount of unlabelled art-related data. We then fine-tune this new pre-trained model with our biography pair dataset manually annotated by a team of professionals in the art industry. With extensive experiments, we demonstrate that our ArtLM achieves 85.6% accuracy and 84.0% F1 score and outperforms other baseline models. We also provide a visualisation and a qualitative analysis of the artist network built from ArtLM's outputs.

♻ ☆ Unlocking the potential of two-point cells for energy-efficient and resilient training of deep nets

Ahsan Adeel, Adewale Adetomi, Khubaib Ahmed, Amir Hussain, Tughrul Arslan, W. A. Phillips

Context-sensitive two-point layer 5 pyramidal cells (L5PCs) were discovered as long ago as 1999. However, the potential of this discovery to provide useful neural computation has yet to be demonstrated. Here we show for the first time how a transformative L5PCs-driven deep neural network (DNN), termed the multisensory cooperative computing (MCC) architecture, can effectively process large amounts of heterogeneous real-world audio-visual (AV) data, using far less energy compared to best available 'point' neuron-driven DNNs. A novel highly-distributed parallel implementation on a Xilinx UltraScale+ MPSoC device estimates energy savings up to 245759 $ \times $ 50000 $\mu$J (i.e., 62% less than the baseline model in a semi-supervised learning setup) where a single synapse consumes $8e^{-5}\mu$J. In a supervised learning setup, the energy-saving can potentially reach up to 1250x less (per feedforward transmission) than the baseline model. The significantly reduced neural activity in MCC leads to inherently fast learning and resilience against sudden neural damage. This remarkable performance in pilot experiments demonstrates the embodied neuromorphic intelligence of our proposed cooperative L5PC that receives input from diverse neighbouring neurons as context to amplify the transmission of most salient and relevant information for onward transmission, from overwhelmingly large multimodal information utilised at the early stages of on-chip training. Our proposed approach opens new cross-disciplinary avenues for future on-chip DNN training implementations and posits a radical shift in current neuromorphic computing paradigms.

♻ ☆ Compositional generalization in semantic parsing with pretrained transformers

A. Emin Orhan

Large-scale pretraining instills large amounts of knowledge in deep neural networks. This, in turn, improves the generalization behavior of these models in downstream tasks. What exactly are the limits to the generalization benefits of large-scale pretraining? Here, we report observations from some simple experiments aimed at addressing this question in the context of two semantic parsing tasks involving natural language, SCAN and COGS. We show that language models pretrained exclusively with non-English corpora, or even with programming language corpora, significantly improve out-of-distribution generalization in these benchmarks, compared with models trained from scratch, even though both benchmarks are English-based. This demonstrates the surprisingly broad transferability of pretrained representations and knowledge. Pretraining with a large-scale protein sequence prediction task, on the other hand, mostly deteriorates the generalization performance in SCAN and COGS, suggesting that pretrained representations do not transfer universally and that there are constraints on the similarity between the pretraining and downstream domains for successful transfer. Finally, we show that larger models are harder to train from scratch and their generalization accuracy is lower when trained up to convergence on the relatively small SCAN and COGS datasets, but the benefits of large-scale pretraining become much clearer with larger models.

comment: v3 adds one reference, adds further discussion, slightly changes formatting

♻ ☆ Missing Data Imputation and Acquisition with Deep Hierarchical Models and Hamiltonian Monte Carlo NeurIPS 2022

Ignacio Peis, Chao Ma, José Miguel Hernández-Lobato

Variational Autoencoders (VAEs) have recently been highly successful at imputing and acquiring heterogeneous missing data. However, within this specific application domain, existing VAE methods are restricted by using only one layer of latent variables and strictly Gaussian posterior approximations. To address these limitations, we present HH-VAEM, a Hierarchical VAE model for mixed-type incomplete data that uses Hamiltonian Monte Carlo with automatic hyper-parameter tuning for improved approximate inference. Our experiments show that HH-VAEM outperforms existing baselines in the tasks of missing data imputation and supervised learning with missing features. Finally, we also present a sampling-based approach for efficiently computing the information gain when missing features are to be acquired with HH-VAEM. Our experiments show that this sampling-based approach is superior to alternatives based on Gaussian approximations.

comment: Published at NeurIPS 2022

♻ ☆ Secure Over-the-Air Computation using Zero-Forced Artificial Noise

Luis Maßny, Antonia Wachter-Zeh

Over-the-air computation has the potential to increase the communication-efficiency of data-dependent distributed wireless systems, but is vulnerable to eavesdropping. We consider over-the-air computation over block-fading additive white Gaussian noise channels in the presence of a passive eavesdropper. The goal is to design a secure over-the-air computation scheme. We propose a scheme that achieves MSE-security against the eavesdropper by employing zero-forced artificial noise, while keeping the distortion at the legitimate receiver small. In contrast to former approaches, the security does not depend on external helper nodes to jam the eavesdropper's received signal. We thoroughly design the system parameters of the scheme, propose an artificial noise design that harnesses unused transmit power for security, and give an explicit construction rule. Our design approach is applicable in both cases, if the eavesdropper's channel coefficients are known and if they are unknown in the signal design. Simulations demonstrate the performance, and show that our noise design outperforms other methods.

♻ ☆ Bit-Metric Decoding Rate in Multi-User MIMO Systems: Theory

K. Pavan Srinath, Jakob Hoydis

Link-adaptation (LA) is one of the most important aspects of wireless communications where the modulation and coding scheme (MCS) used by the transmitter is adapted to the channel conditions in order to meet a certain target error-rate. In a single-user SISO (SU-SISO) system with out-of-cell interference, LA is performed by computing the post-equalization signal-to-interference-noise ratio (SINR) at the receiver. The same technique can be employed in multi-user MIMO (MU-MIMO) receivers that use linear detectors. Another important use of post-equalization SINR is for physical layer (PHY) abstraction, where several PHY blocks like the channel encoder, the detector, and the channel decoder are replaced by an abstraction model in order to speed up system-level simulations. However, for MU-MIMO systems with non-linear receivers, there is no known equivalent of post-equalization SINR which makes both LA and PHY abstraction extremely challenging. This important issue is addressed in this two-part paper. In this part, a metric called the bit-metric decoding rate (BMDR) of a detector, which is the proposed equivalent of post-equalization SINR, is presented. Since BMDR does not have a closed form expression that would enable its instantaneous calculation, a machine-learning approach to predict it is presented along with extensive simulation results.

comment: This is the first part of a two-part paper and has 30 pages and 6 figures. The second part is titled "Bit-Metric Decoding Rate in Multi-User MIMO Systems: Applications". This part has been significantly revised. In particular, we relate BMDR to the mismatched decoding framework that exists in the literature, and have rewritten the claims and proof of the main theorem

♻ ☆ Towards Futuristic Autonomous Experimentation--A Surprise-Reacting Sequential Experiment Policy

Imtiaz Ahmed, Satish Bukkapatnam, Bhaskar Botcha, Yu Ding

An autonomous experimentation platform in manufacturing is supposedly capable of conducting a sequential search for finding suitable manufacturing conditions for advanced materials by itself or even for discovering new materials with minimal human intervention. The core of the intelligent control of such platforms is the policy directing sequential experiments, namely, to decide where to conduct the next experiment based on what has been done thus far. Such policy inevitably trades off exploitation versus exploration and the current practice is under the Bayesian optimization framework using the expected improvement criterion or its variants. We discuss whether it is beneficial to trade off exploitation versus exploration by measuring the element and degree of surprise associated with the immediate past observation. We devise a surprise-reacting policy using two existing surprise metrics, known as the Shannon surprise and Bayesian surprise. Our analysis shows that the surprise-reacting policy appears to be better suited for quickly characterizing the overall landscape of a response surface or a design place under resource constraints. We argue that such capability is much needed for futuristic autonomous experimentation platforms. We do not claim that we have a fully autonomous experimentation platform, but believe that our current effort sheds new lights or provides a different view angle as researchers are racing to elevate the autonomy of various primitive autonomous experimentation systems.

♻ ☆ ReViSe: Remote Vital Signs Measurement Using Smartphone Camera

Donghao Qiao, Amtul Haq Ayesha, Farhana Zulkernine, Raihan Masroor, Nauman Jaffar

We propose an end-to-end framework to measure people's vital signs including Heart Rate (HR), Heart Rate Variability (HRV), Oxygen Saturation (SpO2) and Blood Pressure (BP) based on the rPPG methodology from the video of a user's face captured with a smartphone camera. We extract face landmarks with a deep learning-based neural network model in real-time. Multiple face patches also called Regions-of-Interest (RoIs) are extracted by using the predicted face landmarks. Several filters are applied to reduce the noise from the RoIs in the extracted cardiac signals called Blood Volume Pulse (BVP) signal. The measurements of HR, HRV and SpO2 are validated on two public rPPG datasets namely the TokyoTech rPPG and the Pulse Rate Detection (PURE) datasets, on which our models achieved the following Mean Absolute Errors (MAE): a) for HR, 1.73Beats-Per-Minute (bpm) and 3.95bpm respectively; b) for HRV, 18.55ms and 25.03ms respectively, and c) for SpO2, an MAE of 1.64% on the PURE dataset. We validated our end-to-end rPPG framework, ReViSe, in daily living environment, and thereby created the Video-HR dataset. Our HR estimation model achieved an MAE of 2.49bpm on this dataset. Since no publicly available rPPG datasets existed for BP measurement with face videos, we used a dataset with signals from fingertip sensor to train our deep learning-based BP estimation model and also created our own video dataset, Video-BP. On our Video-BP dataset, our BP estimation model achieved an MAE of 6.7mmHg for Systolic Blood Pressure (SBP), and an MAE of 9.6mmHg for Diastolic Blood Pressure (DBP). ReViSe framework has been validated on datasets with videos recorded in daily living environment as opposed to less noisy laboratory environment as reported by most state-of-the-art techniques.

♻ ☆ Encrypted machine learning of molecular quantum properties

Jan Weinreich, Guido Falk von Rudorff, O. Anatole von Lilienfeld

Large machine learning models with improved predictions have become widely available in the chemical sciences. Unfortunately, these models do not protect the privacy necessary within commercial settings, prohibiting the use of potentially extremely valuable data by others. Encrypting the prediction process can solve this problem by double-blind model evaluation and prohibits the extraction of training or query data. However, contemporary ML models based on fully homomorphic encryption or federated learning are either too expensive for practical use or have to trade higher speed for weaker security. We have implemented secure and computationally feasible encrypted machine learning models using oblivious transfer enabling and secure predictions of molecular quantum properties across chemical compound space. However, we find that encrypted predictions using kernel ridge regression models are a million times more expensive than without encryption. This demonstrates a dire need for a compact machine learning model architecture, including molecular representation and kernel matrix size, that minimizes model evaluation costs.

♻ ☆ Many-Objective Reinforcement Learning for Online Testing of DNN-Enabled Systems

Fitash Ul Haq, Donghwan Shin, Lionel Briand

Deep Neural Networks (DNNs) have been widely used to perform real-world tasks in cyber-physical systems such as Autonomous Driving Systems (ADS). Ensuring the correct behavior of such DNN-Enabled Systems (DES) is a crucial topic. Online testing is one of the promising modes for testing such systems with their application environments (simulated or real) in a closed loop taking into account the continuous interaction between the systems and their environments. However, the environmental variables (e.g., lighting conditions) that might change during the systems' operation in the real world, causing the DES to violate requirements (safety, functional), are often kept constant during the execution of an online test scenario due to the two major challenges: (1) the space of all possible scenarios to explore would become even larger if they changed and (2) there are typically many requirements to test simultaneously. In this paper, we present MORLOT (Many-Objective Reinforcement Learning for Online Testing), a novel online testing approach to address these challenges by combining Reinforcement Learning (RL) and many-objective search. MORLOT leverages RL to incrementally generate sequences of environmental changes while relying on many-objective search to determine the changes so that they are more likely to achieve any of the uncovered objectives. We empirically evaluate MORLOT using CARLA, a high-fidelity simulator widely used for autonomous driving research, integrated with Transfuser, a DNN-enabled ADS for end-to-end driving. The evaluation results show that MORLOT is significantly more effective and efficient than alternatives with a large effect size. In other words, MORLOT is a good option to test DES with dynamically changing environments while accounting for multiple safety requirements.

♻ ☆ On the Sparse DAG Structure Learning Based on Adaptive Lasso

Danru Xu, Erdun Gao, Wei Huang, Andy Song, Mingming Gong

Learning the underlying Bayesian Networks (BNs), represented by directed acyclic graphs (DAGs), of the concerned events from purely-observational data is a crucial part of evidential reasoning. This task remains challenging due to the large and discrete search space. A recent flurry of developments followed NOTEARS[1] recast this combinatorial problem into a continuous optimization problem by leveraging an algebraic equality characterization of acyclicity. However, the continuous optimization methods suffer from obtaining non-spare graphs after the numerical optimization, which leads to the inflexibility to rule out the potentially cycle-inducing edges or false discovery edges with small values. To address this issue, in this paper, we develop a completely data-driven DAG structure learning method without a predefined value to post-threshold small values. We name our method NOTEARS with adaptive Lasso (NOTEARS-AL), which is achieved by applying the adaptive penalty method to ensure the sparsity of the estimated DAG. Moreover, we show that NOTEARS-AL also inherits the oracle properties under some specific conditions. Extensive experiments on both synthetic and a real-world dataset verify the efficacy of the proposed method.

comment: 11 pages, 8 figures

♻ ☆ A3GC-IP: Attention-Oriented Adjacency Adaptive Recurrent Graph Convolutions for Human Pose Estimation from Sparse Inertial Measurements

Patrik Puchert, Timo Ropinski

Conventional methods for human pose estimation either require a high degree of instrumentation, by relying on many inertial measurement units (IMUs), or constraint the recording space, by relying on extrinsic cameras. These deficits are tackled through the approach of human pose estimation from sparse IMU data. We define attention-oriented adjacency adaptive graph convolutional long-short term memory networks (A3GC-LSTM), to tackle human pose estimation based on six IMUs, through incorporating the human body graph structure directly into the network. The A3GC-LSTM combines both spatial and temporal dependency in a single network operation, more memory efficiently than previous approaches. The recurrent graph learning on arbitrarily long sequences is made possible by equipping graph convolutions with adjacency adaptivity, which eliminates the problem of information loss in deep or recurrent graph networks, while it also allows for learning unknown dependencies between the human body joints. To further boost accuracy, a spatial attention formalism is incorporated into the recurrent LSTM cell. With our presented approach, we are able to utilize the inherent graph nature of the human body, and thus can outperform the state of the art for human pose estimation from sparse IMU data.

comment: Preprint, in submission

♻ ☆ A Memetic Algorithm with Reinforcement Learning for Sociotechnical Production Scheduling

Felix Grumbach, Nour Eldin Alaa Badr, Pascal Reusch, Sebastian Trojahn

The following article presents a memetic algorithm with applying deep reinforcement learning (DRL) for solving practically oriented dual resource constrained flexible job shop scheduling problems (DRC-FJSSP). In recent years, there has been extensive research on DRL techniques, but without considering realistic, flexible and human-centered shopfloors. A research gap can be identified in the context of make-to-order oriented discontinuous manufacturing as it is often represented in medium-size companies with high service levels. From practical industry projects in this domain, we recognize requirements to depict flexible machines, human workers and capabilities, setup and processing operations, material arrival times, complex job paths with parallel tasks for bill of material (BOM) manufacturing, sequence-depended setup times and (partially) automated tasks. On the other hand, intensive research has been done on metaheuristics in the context of DRC-FJSSP. However, there is a lack of suitable and generic scheduling methods that can be holistically applied in sociotechnical production and assembly processes. In this paper, we first formulate an extended DRC-FJSSP induced by the practical requirements mentioned. Then we present our proposed hybrid framework with parallel computing for multicriteria optimization. Through numerical experiments with real-world data, we confirm that the framework generates feasible schedules efficiently and reliably. Utilizing DRL instead of random operations leads to better results and outperforms traditional approaches.

comment: This article was submitted to IEEE Access in October 2022. The provisions on copyright specified by IEEE apply: https://www.ieee.org/publications/rights/copyright-policy.html

♻ ☆ A Simple Unified Approach to Testing High-Dimensional Conditional Independences for Categorical and Ordinal Data

Ankur Ankan, Johannes Textor

Conditional independence (CI) tests underlie many approaches to model testing and structure learning in causal inference. Most existing CI tests for categorical and ordinal data stratify the sample by the conditioning variables, perform simple independence tests in each stratum, and combine the results. Unfortunately, the statistical power of this approach degrades rapidly as the number of conditioning variables increases. Here we propose a simple unified CI test for ordinal and categorical data that maintains reasonable calibration and power in high dimensions. We show that our test outperforms existing baselines in model testing and structure learning for dense directed graphical models while being comparable for sparse models. Our approach could be attractive for causal model testing because it is easy to implement, can be used with non-parametric or parametric probability models, has the symmetry property, and has reasonable computational requirements.

♻ ☆ Localising In-Domain Adaptation of Transformer-Based Biomedical Language Models

Tommaso Mario Buonocore, Claudio Crema, Alberto Redolfi, Riccardo Bellazzi, Enea Parimbelli

In the era of digital healthcare, the huge volumes of textual information generated every day in hospitals constitute an essential but underused asset that could be exploited with task-specific, fine-tuned biomedical language representation models, improving patient care and management. For such specialized domains, previous research has shown that fine-tuning models stemming from broad-coverage checkpoints can largely benefit additional training rounds over large-scale in-domain resources. However, these resources are often unreachable for less-resourced languages like Italian, preventing local medical institutions to employ in-domain adaptation. In order to reduce this gap, our work investigates two accessible approaches to derive biomedical language models in languages other than English, taking Italian as a concrete use-case: one based on neural machine translation of English resources, favoring quantity over quality; the other based on a high-grade, narrow-scoped corpus natively written in Italian, thus preferring quality over quantity. Our study shows that data quantity is a harder constraint than data quality for biomedical adaptation, but the concatenation of high-quality data can improve model performance even when dealing with relatively size-limited corpora. The models published from our investigations have the potential to unlock important research opportunities for Italian hospitals and academia. Finally, the set of lessons learned from the study constitutes valuable insights towards a solution to build biomedical language models that are generalizable to other less-resourced languages and different domain settings.

comment: 8 pages, 3 figures

♻ ☆ Sparsely-gated MoE Layers for CNN Interpretability

Svetlana Pavlitskaya, Christian Hubschneider, Lukas Struppek, J. Marius Zöllner

Sparsely-gated Mixture of Expert (MoE) layers have been recently successfully applied for scaling large transformers, especially for language modeling tasks. An intriguing side effect of sparse MoE layers is that they convey inherent interpretability to a model via natural expert specialization. In this work, we apply sparse MoE layers to CNNs for computer vision tasks and analyze the resulting effect on model interpretability. To stabilize MoE training, we present both soft and hard constraint-based approaches. With hard constraints, the weights of certain experts are allowed to become zero, while soft constraints balance the contribution of experts with an additional auxiliary loss. As a result, soft constraints handle expert utilization better and support the expert specialization process, while hard constraints maintain more generalized experts and increase overall model performance. Our findings demonstrate that experts can implicitly focus on individual sub-domains of the input space. For example, experts trained for CIFAR-100 image classification specialize in recognizing different domains such as flowers or animals without previous data clustering. Experiments with RetinaNet and the COCO dataset further indicate that object detection experts can also specialize in detecting objects of distinct sizes.

♻ ☆ A Theoretical Study of The Effects of Adversarial Attacks on Sparse Regression

Deepak Maurya, Jean Honorio

This paper analyzes $\ell_1$ regularized linear regression under the challenging scenario of having only adversarially corrupted data for training. We use the primal-dual witness paradigm to provide provable performance guarantees for the support of the estimated regression parameter vector to match the actual parameter. Our theoretical analysis shows the counter-intuitive result that an adversary can influence sample complexity by corrupting the irrelevant features, i.e., those corresponding to zero coefficients of the regression parameter vector, which, consequently, do not affect the dependent variable. As any adversarially robust algorithm has its limitations, our theoretical analysis identifies the regimes under which the learning algorithm and adversary can dominate over each other. It helps us to analyze these fundamental limits and address critical scientific questions of which parameters (like mutual incoherence, the maximum and minimum eigenvalue of the covariance matrix, and the budget of adversarial perturbation) play a role in the high or low probability of success of the LASSO algorithm. Also, the derived sample complexity is logarithmic with respect to the size of the regression parameter vector, and our theoretical claims are validated by empirical analysis on synthetic and real-world datasets.

comment: first version

♻ ☆ Continual Learning with Distributed Optimization: Does CoCoA Forget?

Martin Hellkvist, Ayça Özçelikkale, Anders Ahlén

We focus on the continual learning problem where the tasks arrive sequentially and the aim is to perform well on the newly arrived task without performance degradation on the previously seen tasks. In contrast to the continual learning literature focusing on the centralized setting, we investigate the distributed estimation framework. We consider the well-established distributed learning algorithm CoCoA. We derive closed form expressions for the iterations for the overparametrized case. We illustrate the convergence and the error performance of the algorithm based on the over/under-parametrization of the problem. Our results show that depending on the problem dimensions and data generation assumptions, CoCoA can perform continual learning over a sequence of tasks, i.e., it can learn a new task without forgetting previously learned tasks, with access only to one task at a time.

♻ ☆ Localizing Anatomical Landmarks in Ocular Images using Zoom-In Attentive Networks

Xiaofeng Lei, Shaohua Li, Xinxing Xu, Huazhu Fu, Yong Liu, Yih-Chung Tham, Yangqin Feng, Mingrui Tan, Yanyu Xu, Jocelyn Hui Lin Goh, Rick Siow Mong Goh, Ching-Yu Cheng

Localizing anatomical landmarks are important tasks in medical image analysis. However, the landmarks to be localized often lack prominent visual features. Their locations are elusive and easily confused with the background, and thus precise localization highly depends on the context formed by their surrounding areas. In addition, the required precision is usually higher than segmentation and object detection tasks. Therefore, localization has its unique challenges different from segmentation or detection. In this paper, we propose a zoom-in attentive network (ZIAN) for anatomical landmark localization in ocular images. First, a coarse-to-fine, or "zoom-in" strategy is utilized to learn the contextualized features in different scales. Then, an attentive fusion module is adopted to aggregate multi-scale features, which consists of 1) a co-attention network with a multiple regions-of-interest (ROIs) scheme that learns complementary features from the multiple ROIs, 2) an attention-based fusion module which integrates the multi-ROIs features and non-ROI features. We evaluated ZIAN on two open challenge tasks, i.e., the fovea localization in fundus images and scleral spur localization in AS-OCT images. Experiments show that ZIAN achieves promising performances and outperforms state-of-the-art localization methods. The source code and trained models of ZIAN are available at https://github.com/leixiaofeng-astar/OMIA9-ZIAN.

♻ ☆ Black-box Error Diagnosis in Deep Neural Networks for Computer Vision: a Survey of Tools

Piero Fraternali, Federico Milani, Rocio Nahime Torres, Niccolò Zangrando

The application of Deep Neural Networks (DNNs) to a broad variety of tasks demands methods for coping with the complex and opaque nature of these architectures. When a gold standard is available, performance assessment treats the DNN as a black box and computes standard metrics based on the comparison of the predictions with the ground truth. A deeper understanding of performances requires going beyond such evaluation metrics to diagnose the model behavior and the prediction errors. This goal can be pursued in two complementary ways. On one side, model interpretation techniques "open the box" and assess the relationship between the input, the inner layers and the output, so as to identify the architecture modules most likely to cause the performance loss. On the other hand, black-box error diagnosis techniques study the correlation between the model response and some properties of the input not used for training, so as to identify the features of the inputs that make the model fail. Both approaches give hints on how to improve the architecture and/or the training process. This paper focuses on the application of DNNs to Computer Vision (CV) tasks and presents a survey of the tools that support the black-box performance diagnosis paradigm. It illustrates the features and gaps of the current proposals, discusses the relevant research directions and provides a brief overview of the diagnosis tools in sectors other than CV.

comment: Published in Springer Neural Computing and Applications, https://link.springer.com/article/10.1007/s00521-022-08100-9

♻ ☆ Yes We Care! -- Certification for Machine Learning Methods through the Care Label Framework

Katharina Morik, Helena Kotthaus, Raphael Fischer, Sascha Mücke, Matthias Jakobs, Nico Piatkowski, Andreas Pauly, Lukas Heppe, Danny Heinrich

Machine learning applications have become ubiquitous. Their applications range from embedded control in production machines over process optimization in diverse areas (e.g., traffic, finance, sciences) to direct user interactions like advertising and recommendations. This has led to an increased effort of making machine learning trustworthy. Explainable and fair AI have already matured. They address the knowledgeable user and the application engineer. However, there are users that want to deploy a learned model in a similar way as their washing machine. These stakeholders do not want to spend time in understanding the model, but want to rely on guaranteed properties. What are the relevant properties? How can they be expressed to the stakeholder without presupposing machine learning knowledge? How can they be guaranteed for a certain implementation of a machine learning model? These questions move far beyond the current state of the art and we want to address them here. We propose a unified framework that certifies learning methods via care labels. They are easy to understand and draw inspiration from well-known certificates like textile labels or property cards of electronic devices. Our framework considers both, the machine learning theory and a given implementation. We test the implementation's compliance with theoretical properties and bounds.

♻ ☆ Multiscale Graph Neural Networks for Protein Residue Contact Map Prediction

Kuang Liu, Rajiv K. Kalia, Xinlian Liu, Aiichiro Nakano, Ken-ichi Nomura, Priya Vashishta, Rafael Zamora-Resendizc

Machine learning (ML) is revolutionizing protein structural analysis, including an important subproblem of predicting protein residue contact maps, i.e., which amino-acid residues are in close spatial proximity given the amino-acid sequence of a protein. Despite recent progresses in ML-based protein contact prediction, predicting contacts with a wide range of distances (commonly classified into short-, medium- and long-range contacts) remains a challenge. Here, we propose a multiscale graph neural network (GNN) based approach taking a cue from multiscale physics simulations, in which a standard pipeline involving a recurrent neural network (RNN) is augmented with three GNNs to refine predictive capability for short-, medium- and long-range residue contacts, respectively. Test results on the ProteinNet dataset show improved accuracy for contacts of all ranges using the proposed multiscale RNN+GNN approach over the conventional approach, including the most challenging case of long-range contact prediction.

♻ ☆ Theoretically Motivated Data Augmentation and Regularization for Portfolio Construction

Liu Ziyin, Kentaro Minami, Kentaro Imajo

The task we consider is portfolio construction in a speculative market, a fundamental problem in modern finance. While various empirical works now exist to explore deep learning in finance, the theory side is almost non-existent. In this work, we focus on developing a theoretical framework for understanding the use of data augmentation for deep-learning-based approaches to quantitative finance. The proposed theory clarifies the role and necessity of data augmentation for finance; moreover, our theory implies that a simple algorithm of injecting a random noise of strength $\sqrt{|r_{t-1}|}$ to the observed return $r_{t}$ is better than not injecting any noise and a few other financially irrelevant data augmentation techniques.

comment: The full version of our work published at 3rd ACM International Conference on AI in Finance (ICAIF'22)

♻ ☆ Goal-Conditioned Q-Learning as Knowledge Distillation AAAI 2023

Alexander Levine, Soheil Feizi

Many applications of reinforcement learning can be formalized as goal-conditioned environments, where, in each episode, there is a "goal" that affects the rewards obtained during that episode but does not affect the dynamics. Various techniques have been proposed to improve performance in goal-conditioned environments, such as automatic curriculum generation and goal relabeling. In this work, we explore a connection between off-policy reinforcement learning in goal-conditioned settings and knowledge distillation. In particular: the current Q-value function and the target Q-value estimate are both functions of the goal, and we would like to train the Q-value function to match its target for all goals. We therefore apply Gradient-Based Attention Transfer (Zagoruyko and Komodakis 2017), a knowledge distillation technique, to the Q-function update. We empirically show that this can improve the performance of goal-conditioned off-policy reinforcement learning when the space of goals is high-dimensional. We also show that this technique can be adapted to allow for efficient learning in the case of multiple simultaneous sparse goals, where the agent can attain a reward by achieving any one of a large set of objectives, all specified at test time. Finally, to provide theoretical support, we give examples of classes of environments where (under some assumptions) standard off-policy algorithms such as DDPG require at least O(d^2) replay buffer transitions to learn an optimal policy, while our proposed technique requires only O(d) transitions, where d is the dimensionality of the goal and state space. Code is available at https://github.com/alevine0/ReenGAGE.

comment: AAAI 2023 Accepted paper

♻ ☆ Dynamic Budget Throttling in Repeated Second-Price Auctions

Zhaohua Chen, Chang Wang, Qian Wang, Yuqi Pan, Zhuming Shi, Zheng Cai, Yukun Ren, Zhihua Zhu, Xiaotie Deng

In today's online advertising markets, an important demand for an advertiser (buyer) is to control her total expenditure within a time span under some budget. Among all budget control approaches, throttling stands out as a popular one, where the buyer participates in only a part of auctions. This paper gives a theoretical panorama of a single buyer's dynamic budget throttling process in repeated second-price auctions, which is lacking in the literature. We first establish a lower bound on the regret and an upper bound on the asymptotic competitive ratio for any throttling algorithm, respectively, on whether the buyer's values are stochastic or adversarial. Second, on the algorithmic side, we consider two different information structures, with increasing difficulty in learning the stochastic distribution of the highest competing bid. We further propose the OGD-CB algorithm, which is oblivious to stochastic or adversarial values and has asymptotically equal results under these two information structures. Specifically, with stochastic values, we demonstrate that this algorithm guarantees a near-optimal expected regret. When values are adversarial, we prove that the proposed algorithm reaches the upper bound on the asymptotic competitive ratio. At last, we compare throttling with pacing, another widely adopted budget control method, in repeated second-price auctions. In the stochastic case, we illustrate that pacing is generally better than throttling for the buyer, which is an extension of known results that pacing is asymptotically optimal in this scenario. However, in the adversarial case, we give an exciting result indicating that throttling is the asymptotically optimal dynamic bidding strategy. Our results fill the gaps in the theoretical research of throttling in repeated auctions and comprehensively reveal the ability of this popular budget-smoothing strategy.

comment: 45 pages, 1 figure, 1 table

♻ ☆ Robust Path Selection in Software-defined WANs using Deep Reinforcement Learning

Shahrooz Pouryousef, Lixin Gao, Don Towsley

In the context of an efficient network traffic engineering process where the network continuously measures a new traffic matrix and updates the set of paths in the network, an automated process is required to quickly and efficiently identify when and what set of paths should be used. Unfortunately, the burden of finding the optimal solution for the network updating process in each given time interval is high since the computation complexity of optimization approaches using linear programming increases significantly as the size of the network increases. In this paper, we use deep reinforcement learning to derive a data-driven algorithm that does the path selection in the network considering the overhead of route computation and path updates. Our proposed scheme leverages information about past network behavior to identify a set of robust paths to be used for multiple future time intervals to avoid the overhead of updating the forwarding behavior of routers frequently. We compare the results of our approach to other traffic engineering solutions through extensive simulations across real network topologies. Our results demonstrate that our scheme fares well by a factor of 40% with respect to reducing link utilization compared to traditional TE schemes such as ECMP. Our scheme provides a slightly higher link utilization (around 25%) compared to schemes that only minimize link utilization and do not care about path updating overhead.

♻ ☆ SlimFL: Federated Learning with Superposition Coding over Slimmable Neural Networks

Won Joon Yun, Yunseok Kwak, Hankyul Baek, Soyi Jung, Mingyue Ji, Mehdi Bennis, Jihong Park, Joongheon Kim

Federated learning (FL) is a key enabler for efficient communication and computing, leveraging devices' distributed computing capabilities. However, applying FL in practice is challenging due to the local devices' heterogeneous energy, wireless channel conditions, and non-independently and identically distributed (non-IID) data distributions. To cope with these issues, this paper proposes a novel learning framework by integrating FL and width-adjustable slimmable neural networks (SNN). Integrating FL with SNNs is challenging due to time-varying channel conditions and data distributions. In addition, existing multi-width SNN training algorithms are sensitive to the data distributions across devices, which makes SNN ill-suited for FL. Motivated by this, we propose a communication and energy-efficient SNN-based FL (named SlimFL) that jointly utilizes superposition coding (SC) for global model aggregation and superposition training (ST) for updating local models. By applying SC, SlimFL exchanges the superposition of multiple-width configurations decoded as many times as possible for a given communication throughput. Leveraging ST, SlimFL aligns the forward propagation of different width configurations while avoiding inter-width interference during backpropagation. We formally prove the convergence of SlimFL. The result reveals that SlimFL is not only communication-efficient but also deals with non-IID data distributions and poor channel conditions, which is also corroborated by data-intensive simulations.

comment: arXiv admin note: text overlap with arXiv:2112.02543

♻ ☆ Generalized Stable Weights via Neural Gibbs Density

Yoshiaki Kitazawa

We present a generalized balancing method -- stable weights via Neural Gibbs Density -- fully available for estimating causal effects for an arbitrary mixture of discrete and continuous interventions. Our weights are trainable through back-propagation and can be obtained with neural network algorithms. In addition, we also provide a method to measure the performance of our weights by estimating the mutual information for the balanced distribution. Our method is easy to implement with any present deep learning libraries, and the weights from it can be used in most state-of-art supervised algorithms.

♻ ☆ Dataset Condensation with Distribution Matching

Bo Zhao, Hakan Bilen

Computational cost of training state-of-the-art deep models in many learning problems is rapidly increasing due to more sophisticated models and larger datasets. A recent promising direction for reducing training cost is dataset condensation that aims to replace the original large training set with a significantly smaller learned synthetic set while preserving the original information. While training deep models on the small set of condensed images can be extremely fast, their synthesis remains computationally expensive due to the complex bi-level optimization and second-order derivative computation. In this work, we propose a simple yet effective method that synthesizes condensed images by matching feature distributions of the synthetic and original training images in many sampled embedding spaces. Our method significantly reduces the synthesis cost while achieving comparable or better performance. Thanks to its efficiency, we apply our method to more realistic and larger datasets with sophisticated neural architectures and obtain a significant performance boost. We also show promising practical benefits of our method in continual learning and neural architecture search.

♻ ☆ Directed Acyclic Graph Factorization Machines for CTR Prediction via Knowledge Distillation

Zhen Tian, Ting Bai, Zibin Zhang, Zhiyuan Xu, Kangyi Lin, Ji-Rong Wen, Wayne Xin Zhao

With the growth of high-dimensional sparse data in web-scale recommender systems, the computational cost to learn high-order feature interaction in CTR prediction task largely increases, which limits the use of high-order interaction models in real industrial applications. Some recent knowledge distillation based methods transfer knowledge from complex teacher models to shallow student models for accelerating the online model inference. However, they suffer from the degradation of model accuracy in knowledge distillation process. It is challenging to balance the efficiency and effectiveness of the shallow student models. To address this problem, we propose a Directed Acyclic Graph Factorization Machine (KD-DAGFM) to learn the high-order feature interactions from existing complex interaction models for CTR prediction via Knowledge Distillation. The proposed lightweight student model DAGFM can learn arbitrary explicit feature interactions from teacher networks, which achieves approximately lossless performance and is proved by a dynamic programming algorithm. Besides, an improved general model KD-DAGFM+ is shown to be effective in distilling both explicit and implicit feature interactions from any complex teacher model. Extensive experiments are conducted on four real-world datasets, including a large-scale industrial dataset from WeChat platform with billions of feature dimensions. KD-DAGFM achieves the best performance with less than 21.5% FLOPs of the state-of-the-art method on both online and offline experiments, showing the superiority of DAGFM to deal with the industrial scale data in CTR prediction task. Our implementation code is available at: https://github.com/RUCAIBox/DAGFM.

♻ ☆ Self-Supervised Contrastive Representation Learning for 3D Mesh Segmentation AAAI 2023

Ayaan Haque, Hankyu Moon, Heng Hao, Sima Didari, Jae Oh Woo, Patrick Bangert

3D deep learning is a growing field of interest due to the vast amount of information stored in 3D formats. Triangular meshes are an efficient representation for irregular, non-uniform 3D objects. However, meshes are often challenging to annotate due to their high geometrical complexity. Specifically, creating segmentation masks for meshes is tedious and time-consuming. Therefore, it is desirable to train segmentation networks with limited-labeled data. Self-supervised learning (SSL), a form of unsupervised representation learning, is a growing alternative to fully-supervised learning which can decrease the burden of supervision for training. We propose SSL-MeshCNN, a self-supervised contrastive learning method for pre-training CNNs for mesh segmentation. We take inspiration from traditional contrastive learning frameworks to design a novel contrastive learning algorithm specifically for meshes. Our preliminary experiments show promising results in reducing the heavy labeled data requirement needed for mesh segmentation by at least 33%.

comment: AAAI 2023

☆ Generative Colorization of Structured Mobile Web Pages WACV 2023

Kotaro Kikuchi, Naoto Inoue, Mayu Otani, Edgar Simo-Serra, Kota Yamaguchi

Color is a critical design factor for web pages, affecting important factors such as viewer emotions and the overall trust and satisfaction of a website. Effective coloring requires design knowledge and expertise, but if this process could be automated through data-driven modeling, efficient exploration and alternative workflows would be possible. However, this direction remains underexplored due to the lack of a formalization of the web page colorization problem, datasets, and evaluation protocols. In this work, we propose a new dataset consisting of e-commerce mobile web pages in a tractable format, which are created by simplifying the pages and extracting canonical color styles with a common web browser. The web page colorization problem is then formalized as a task of estimating plausible color styles for a given web page content with a given hierarchical structure of the elements. We present several Transformer-based methods that are adapted to this task by prepending structural message passing to capture hierarchical relationships between elements. Experimental results, including a quantitative evaluation designed for this task, demonstrate the advantages of our methods over statistical and image colorization methods. The code is available at https://github.com/CyberAgentAILab/webcolor.

comment: Accepted to WACV 2023

☆ Multi-queue Momentum Contrast for Microvideo-Product Retrieval WSDM '23

Yali Du, Yinwei Wei, Wei Ji, Fan Liu, Xin Luo, Liqiang Nie

The booming development and huge market of micro-videos bring new e-commerce channels for merchants. Currently, more micro-video publishers prefer to embed relevant ads into their micro-videos, which not only provides them with business income but helps the audiences to discover their interesting products. However, due to the micro-video recording by unprofessional equipment, involving various topics and including multiple modalities, it is challenging to locate the products related to micro-videos efficiently, appropriately, and accurately. We formulate the microvideo-product retrieval task, which is the first attempt to explore the retrieval between the multi-modal and multi-modal instances. A novel approach named Multi-Queue Momentum Contrast (MQMC) network is proposed for bidirectional retrieval, consisting of the uni-modal feature and multi-modal instance representation learning. Moreover, a discriminative selection strategy with a multi-queue is used to distinguish the importance of different negatives based on their categories. We collect two large-scale microvideo-product datasets (MVS and MVS-large) for evaluation and manually construct the hierarchical category ontology, which covers sundry products in daily life. Extensive experiments show that MQMC outperforms the state-of-the-art baselines. Our replication package (including code, dataset, etc.) is publicly available at https://github.com/duyali2000/MQMC.

comment: Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining (WSDM '23), February 27-March 3, 2023, Singapore, Singapore

☆ Generalized Decoding for Pixel, Image, and Language

Xueyan Zou, Zi-Yi Dou, Jianwei Yang, Zhe Gan, Linjie Li, Chunyuan Li, Xiyang Dai, Harkirat Behl, Jianfeng Wang, Lu Yuan, Nanyun Peng, Lijuan Wang, Yong Jae Lee, Jianfeng Gao

We present X-Decoder, a generalized decoding model that can predict pixel-level segmentation and language tokens seamlessly. X-Decodert takes as input two types of queries: (i) generic non-semantic queries and (ii) semantic queries induced from text inputs, to decode different pixel-level and token-level outputs in the same semantic space. With such a novel design, X-Decoder is the first work that provides a unified way to support all types of image segmentation and a variety of vision-language (VL) tasks. Further, our design enables seamless interactions across tasks at different granularities and brings mutual benefits by learning a common and rich pixel-level visual-semantic understanding space, without any pseudo-labeling. After pretraining on a mixed set of a limited amount of segmentation data and millions of image-text pairs, X-Decoder exhibits strong transferability to a wide range of downstream tasks in both zero-shot and finetuning settings. Notably, it achieves (1) state-of-the-art results on open-vocabulary segmentation and referring segmentation on eight datasets; (2) better or competitive finetuned performance to other generalist and specialist models on segmentation and VL tasks; and (3) flexibility for efficient finetuning and novel task composition (e.g., referring captioning and image editing). Code, demo, video, and visualization are available at https://x-decoder-vl.github.io.

comment: https://x-decoder-vl.github.io

☆ Contrastive Language-Vision AI Models Pretrained on Web-Scraped Multimodal Data Exhibit Sexual Objectification Bias

Robert Wolfe, Yiwei Yang, Bill Howe, Aylin Caliskan

Nine language-vision AI models trained on web scrapes with the Contrastive Language-Image Pretraining (CLIP) objective are evaluated for evidence of a bias studied by psychologists: the sexual objectification of girls and women, which occurs when a person's human characteristics are disregarded and the person is treated as a body or a collection of body parts. A first experiment uses standardized images of women from the Sexual OBjectification and EMotion Database, and finds that, commensurate with prior research in psychology, human characteristics are disassociated from images of objectified women: the model's recognition of emotional state is mediated by whether the subject is fully or partially clothed. Embedding association tests (EATs) return significant effect sizes for both anger (d >.8) and sadness (d >.5). A second experiment measures the effect in a representative application: an automatic image captioner (Antarctic Captions) includes words denoting emotion less than 50% as often for images of partially clothed women than for images of fully clothed women. A third experiment finds that images of female professionals (scientists, doctors, executives) are likely to be associated with sexual descriptions relative to images of male professionals. A fourth experiment shows that a prompt of "a [age] year old girl" generates sexualized images (as determined by an NSFW classifier) up to 73% of the time for VQGAN-CLIP (age 17), and up to 40% of the time for Stable Diffusion (ages 14 and 18); the corresponding rate for boys never surpasses 9%. The evidence indicates that language-vision AI models trained on automatically collected web scrapes learn biases of sexual objectification, which propagate to downstream applications.

comment: 15 pages, 2 figures

☆ Improving Narrative Relationship Embeddings by Training with Additional Inverse-Relationship Constraints

Mikolaj Figurski

We consider the problem of embedding character-entity relationships from the reduced semantic space of narratives, proposing and evaluating the assumption that these relationships hold under a reflection operation. We analyze this assumption and compare the approach to a baseline state-of-the-art model with a unique evaluation that simulates efficacy on a downstream clustering task with human-created labels. Although our model creates clusters that achieve Silhouette scores of -.084, outperforming the baseline -.227, our analysis reveals that the models approach the task much differently and perform well on very different examples. We conclude that our assumption might be useful for specific types of data and should be evaluated on a wider range of tasks.

☆ Entropy- and Distance-Based Predictors From GPT-2 Attention Patterns Predict Reading Times Over and Above GPT-2 Surprisal EMNLP 2022

Byung-Doh Oh, William Schuler

Transformer-based large language models are trained to make predictions about the next word by aggregating representations of previous tokens through their self-attention mechanism. In the field of cognitive modeling, such attention patterns have recently been interpreted as embodying the process of cue-based retrieval, in which attention over multiple targets is taken to generate interference and latency during retrieval. Under this framework, this work first defines an entropy-based predictor that quantifies the diffuseness of self-attention, as well as distance-based predictors that capture the incremental change in attention patterns across timesteps. Moreover, following recent studies that question the informativeness of attention weights, we also experiment with alternative methods for incorporating vector norms into attention weights. Regression experiments using predictors calculated from the GPT-2 language model show that these predictors deliver a substantially better fit to held-out self-paced reading and eye-tracking data over a rigorous baseline including GPT-2 surprisal. Additionally, the distance-based predictors generally demonstrated higher predictive power, with effect sizes of up to 6.59 ms per standard deviation on self-paced reading times (compared to 2.82 ms for surprisal) and 1.05 ms per standard deviation on eye-gaze durations (compared to 3.81 ms for surprisal).

comment: EMNLP 2022

☆ Universal versus system-specific features of punctuation usage patterns in~major Western~languages

Tomasz Stanisz, Stanislaw Drozdz, Jaroslaw Kwapien

The celebrated proverb that "speech is silver, silence is golden" has a long multinational history and multiple specific meanings. In written texts punctuation can in fact be considered one of its manifestations. Indeed, the virtue of effectively speaking and writing involves - often decisively - the capacity to apply the properly placed breaks. In the present study, based on a large corpus of world-famous and representative literary texts in seven major Western languages, it is shown that the distribution of intervals between consecutive punctuation marks in almost all texts can universally be characterised by only two parameters of the discrete Weibull distribution which can be given an intuitive interpretation in terms of the so-called hazard function. The values of these two parameters tend to be language-specific, however, and even appear to navigate translations. The properties of the computed hazard functions indicate that among the studied languages, English turns out to be the least constrained by the necessity to place a consecutive punctuation mark to partition a sequence of words. This may suggest that when compared to other studied languages, English is more flexible, in the sense of allowing longer uninterrupted sequences of words. Spanish reveals similar tendency to only a bit lesser extent.

☆ Computer says "No": The Case Against Empathetic Conversational AI

Alba Curry, Amanda Cercas Curry

Emotions are an integral part of human cognition and they guide not only our understanding of the world but also our actions within it. As such, whether we soothe or flame an emotion is not inconsequential. Recent work in conversational AI has focused on responding empathetically to users, validating and soothing their emotions without a real basis. This AI-aided emotional regulation can have negative consequences for users and society, tending towards a one-noted happiness defined as only the absence of "negative" emotions. We argue that we must carefully consider whether and how to respond to users' emotions.

☆ Parallel Context Windows Improve In-Context Learning of Large Language Models

Nir Ratner, Yoav Levine, Yonatan Belinkov, Ori Ram, Omri Abend, Ehud Karpas, Amnon Shashua, Kevin Leyton-Brown, Yoav Shoham

For applications that require processing large amounts of text at inference time, Large Language Models (LLMs) are handicapped by their limited context windows, which are typically 2048 tokens. In-context learning, an emergent phenomenon in LLMs in sizes above a certain parameter threshold, constitutes one significant example because it can only leverage training examples that fit into the context window. Existing efforts to address the context window limitation involve training specialized architectures, which tend to be smaller than the sizes in which in-context learning manifests due to the memory footprint of processing long texts. We present Parallel Context Windows (PCW), a method that alleviates the context window restriction for any off-the-shelf LLM without further training. The key to the approach is to carve a long context into chunks (``windows'') that fit within the architecture, restrict the attention mechanism to apply only within each window, and re-use the positional embeddings among the windows. We test the PCW approach on in-context learning with models that range in size between 750 million and 178 billion parameters, and show substantial improvements for tasks with diverse input and output spaces. Our results motivate further investigation of Parallel Context Windows as a method for applying off-the-shelf LLMs in other settings that require long text sequences.

☆ Critic-Guided Decoding for Controlled Text Generation

Minbeom Kim, Hwanhee Lee, Kang Min Yoo, Joonsuk Park, Hwaran Lee, Kyomin Jung

Steering language generation towards objectives or away from undesired content has been a long-standing goal in utilizing language models (LM). Recent work has demonstrated reinforcement learning and weighted decoding as effective approaches to achieve a higher level of language control and quality with pros and cons. In this work, we propose a novel critic decoding method for controlled language generation (CriticControl) that combines the strengths of reinforcement learning and weighted decoding. Specifically, we adopt the actor-critic framework to train an LM-steering critic from non-differentiable reward models. And similar to weighted decoding, our method freezes the language model and manipulates the output token distribution using called critic, improving training efficiency and stability. Evaluation of our method on three controlled generation tasks, namely topic control, sentiment control, and detoxification, shows that our approach generates more coherent and well-controlled texts than previous methods. In addition, CriticControl demonstrates superior generalization ability in zero-shot settings. Human evaluation studies also corroborate our findings.

comment: 11 pages, 6 figures

☆ Esports Data-to-commentary Generation on Large-scale Data-to-text Dataset

Zihan Wang, Naoki Yoshinaga

Esports, a sports competition using video games, has become one of the most important sporting events in recent years. Although the amount of esports data is increasing than ever, only a small fraction of those data accompanies text commentaries for the audience to retrieve and understand the plays. Therefore, in this study, we introduce a task of generating game commentaries from structured data records to address the problem. We first build a large-scale esports data-to-text dataset using structured data and commentaries from a popular esports game, League of Legends. On this dataset, we devise several data preprocessing methods including linearization and data splitting to augment its quality. We then introduce several baseline encoder-decoder models and propose a hierarchical model to generate game commentaries. Considering the characteristics of esports commentaries, we design evaluation metrics including three aspects of the output: correctness, fluency, and strategic depth. Experimental results on our large-scale esports dataset confirmed the advantage of the hierarchical model, and the results revealed several challenges of this novel task.

☆ Resolving Indirect Referring Expressions for Entity Selection

Mohammad Javad Hosseini, Filip Radlinski, Silvia Pareti, Annie Louis

Recent advances in language modeling have enabled new conversational systems. In particular, it is often desirable for people to make choices among specified options when using such systems. We address the problem of reference resolution, when people use natural expressions to choose between real world entities. For example, given the choice `Should we make a Simnel cake or a Pandan cake?' a natural response from a non-expert may be indirect: `let's make the green one'. Reference resolution has been little studied with natural expressions, thus robustly understanding such language has large potential for improving naturalness in dialog, recommendation, and search systems. We create AltEntities (Alternative Entities), a new public dataset of entity pairs and utterances, and develop models for the disambiguation problem. Consisting of 42K indirect referring expressions across three domains, it enables for the first time the study of how large language models can be adapted to this task. We find they achieve 82%-87% accuracy in realistic settings, which while reasonable also invites further advances.

☆ SPT: Semi-Parametric Prompt Tuning for Multitask Prompted Learning

M Saiful Bari, Aston Zhang, Shuai Zheng, Xingjian Shi, Yi Zhu, Shafiq Joty, Mu Li

Pre-trained large language models can efficiently interpolate human-written prompts in a natural way. Multitask prompted learning can help generalization through a diverse set of tasks at once, thus enhancing the potential for more effective downstream fine-tuning. To perform efficient multitask-inference in the same batch, parameter-efficient fine-tuning methods such as prompt tuning have been proposed. However, the existing prompt tuning methods may lack generalization. We propose SPT, a semi-parametric prompt tuning method for multitask prompted learning. The novel component of SPT is a memory bank from where memory prompts are retrieved based on discrete prompts. Extensive experiments, such as (i) fine-tuning a full language model with SPT on 31 different tasks from 8 different domains and evaluating zero-shot generalization on 9 heldout datasets under 5 NLP task categories and (ii) pretraining SPT on the GLUE datasets and evaluating fine-tuning on the SuperGLUE datasets, demonstrate effectiveness of SPT.

☆ Language Models as Inductive Reasoners

Zonglin Yang, Li Dong, Xinya Du, Hao Cheng, Erik Cambria, Xiaodong Liu, Jianfeng Gao, Furu Wei

Inductive reasoning is a core component of human intelligence. In the past research of inductive reasoning within computer science, logic language is used as representations of knowledge (facts and rules, more specifically). However, logic language can cause systematic problems for inductive reasoning such as disability of handling raw input such as natural language, sensitiveness to mislabeled data, and incapacity to handle ambiguous input. To this end, we propose a new task, which is to induce natural language rules from natural language facts, and create a dataset termed DEER containing 1.2k rule-fact pairs for the task, where rules and facts are written in natural language. New automatic metrics are also proposed and analysed for the evaluation of this task. With DEER, we investigate a modern approach for inductive reasoning where we use natural language as representation for knowledge instead of logic language and use pretrained language models as ''reasoners''. Moreover, we provide the first and comprehensive analysis of how well pretrained language models can induce natural language rules from natural language facts. We also propose a new framework drawing insights from philosophy literature for this task, which we show in the experiment section that surpasses baselines in both automatic and human evaluations.

☆ RECAP: Retrieval Augmented Music Captioner

Zihao He, Weituo Hao, Xuchen Song

With the prevalence of stream media platforms serving music search and recommendation, interpreting music by understanding audio and lyrics interactively has become an important and challenging task. However, many previous works focus on refining individual components of encoder-decoder architecture mapping music to caption tokens, ignoring the potential usage of audio and lyrics correspondence. In this paper, we propose to explicitly learn the multi-modal alignment with retrieval augmentation by contrastive learning. By learning audio-lyrics correspondence, the model is guided to learn better cross-modal attention weights, thus generating high-quality caption words. We provide both theoretical and empirical results that demonstrate the advantage of the proposed method.

☆ Training language models for deeper understanding improves brain alignment

Khai Loong Aw, Mariya Toneva

Building systems that achieve a deeper understanding of language is one of the central goals of natural language processing (NLP). Towards this goal, recent works have begun to train language models on narrative datasets which require extracting the most critical information by integrating across long contexts. However, it is still an open question whether these models are learning a deeper understanding of the text, or if the models are simply learning a heuristic to complete the task. This work investigates this further by turning to the one language processing system that truly understands complex language: the human brain. We show that training language models for deeper narrative understanding results in richer representations that have improved alignment to human brain activity. We further find that the improvements in brain alignment are larger for character names than for other discourse features, which indicates that these models are learning important narrative elements. Taken together, these results suggest that this type of training can indeed lead to deeper language understanding. These findings have consequences both for cognitive neuroscience by revealing some of the significant factors behind brain-NLP alignment, and for NLP by highlighting that understanding of long-range context can be improved beyond language modeling.

comment: 37 pages, 42 figures

☆ A Survey of Mix-based Data Augmentation: Taxonomy, Methods, Applications, and Explainability

Chengtai Cao, Fan Zhou, Yurou Dai, Jianping Wang

Data augmentation (DA) is indispensable in modern machine learning and deep neural networks. The basic idea of DA is to construct new training data to improve the model's generalization by adding slightly disturbed versions of existing data or synthesizing new data. In this work, we review a small but essential subset of DA -- Mix-based Data Augmentation (MixDA) that generates novel samples by mixing multiple examples. Unlike conventional DA approaches based on a single-sample operation or requiring domain knowledge, MixDA is more general in creating a broad spectrum of new data and has received increasing attention in the community. We begin with proposing a new taxonomy classifying MixDA into, Mixup-based, Cutmix-based, and hybrid approaches according to a hierarchical view of the data mix. Various MixDA techniques are then comprehensively reviewed in a more fine-grained way. Owing to its generalization, MixDA has penetrated a variety of applications which are also completely reviewed in this work. We also examine why MixDA works from different aspects of improving model performance, generalization, and calibration while explaining the model behavior based on the properties of MixDA. Finally, we recapitulate the critical findings and fundamental challenges of current MixDA studies, and outline the potential directions for future works. Different from previous related works that summarize the DA approaches in a specific domain (e.g., images or natural language processing) or only review a part of MixDA studies, we are the first to provide a systematical survey of MixDA in terms of its taxonomy, methodology, applications, and explainability. This work can serve as a roadmap to MixDA techniques and application reviews while providing promising directions for researchers interested in this exciting area.

☆ Cross-Linguistic Syntactic Difference in Multilingual BERT: How Good is It and How Does It Affect Transfer? EMNLP 2022

Ningyu Xu, Tao Gui, Ruotian Ma, Qi Zhang, Jingting Ye, Menghan Zhang, Xuanjing Huang

Multilingual BERT (mBERT) has demonstrated considerable cross-lingual syntactic ability, whereby it enables effective zero-shot cross-lingual transfer of syntactic knowledge. The transfer is more successful between some languages, but it is not well understood what leads to this variation and whether it fairly reflects difference between languages. In this work, we investigate the distributions of grammatical relations induced from mBERT in the context of 24 typologically different languages. We demonstrate that the distance between the distributions of different languages is highly consistent with the syntactic difference in terms of linguistic formalisms. Such difference learnt via self-supervision plays a crucial role in the zero-shot transfer performance and can be predicted by variation in morphosyntactic properties between languages. These results suggest that mBERT properly encodes languages in a way consistent with linguistic diversity and provide insights into the mechanism of cross-lingual transfer.

comment: EMNLP 2022

☆ Prompt-Augmented Linear Probing: Scaling Beyond The Limit of Few-shot In-Context Learners AAAI 2023

Hyunsoo Cho, Hyuhng Joon Kim, Junyeob Kim, Sang-Woo Lee, Sang-goo Lee, Kang Min Yoo, Taeuk Kim

Through in-context learning (ICL), large-scale language models are effective few-shot learners without additional model fine-tuning. However, the ICL performance does not scale well with the number of available training samples as it is limited by the inherent input length constraint of the underlying language model. Meanwhile, many studies have revealed that language models are also powerful feature extractors, allowing them to be utilized in a black-box manner and enabling the linear probing paradigm, where lightweight discriminators are trained on top of the pre-extracted input representations. This paper proposes prompt-augmented linear probing (PALP), a hybrid of linear probing and ICL, which leverages the best of both worlds. PALP inherits the scalability of linear probing and the capability of enforcing language models to derive more meaningful representations via tailoring input into a more conceivable form. Throughout in-depth investigations on various datasets, we verified that PALP significantly enhances the input representations closing the gap between ICL in the data-hungry scenario and fine-tuning in the data-abundant scenario with little training overhead, potentially making PALP a strong alternative in a black-box scenario.

comment: AAAI 2023

☆ Generating Multiple-Length Summaries via Reinforcement Learning for Unsupervised Sentence Summarization EMNLP 2022

Dongmin Hyun, Xiting Wang, Chanyoung Park, Xing Xie, Hwanjo Yu

Sentence summarization shortens given texts while maintaining core contents of the texts. Unsupervised approaches have been studied to summarize texts without human-written summaries. However, recent unsupervised models are extractive, which remove words from texts and thus they are less flexible than abstractive summarization. In this work, we devise an abstractive model based on reinforcement learning without ground-truth summaries. We formulate the unsupervised summarization based on the Markov decision process with rewards representing the summary quality. To further enhance the summary quality, we develop a multi-summary learning mechanism that generates multiple summaries with varying lengths for a given text, while making the summaries mutually enhance each other. Experimental results show that the proposed model substantially outperforms both abstractive and extractive models, yet frequently generating new words not contained in input texts.

comment: Findings of EMNLP 2022

☆ End-to-End Automatic Speech Recognition model for the Sudanese Dialect

Ayman Mansour, Wafaa F. Mukhtar

Designing a natural voice interface rely mostly on Speech recognition for interaction between human and their modern digital life equipment. In addition, speech recognition narrows the gap between monolingual individuals to better exchange communication. However, the field lacks wide support for several universal languages and their dialects, while most of the daily conversations are carried out using them. This paper comes to inspect the viability of designing an Automatic Speech Recognition model for the Sudanese dialect, which is one of the Arabic Language dialects, and its complexity is a product of historical and social conditions unique to its speakers. This condition is reflected in both the form and content of the dialect, so this paper gives an overview of the Sudanese dialect and the tasks of collecting represented resources and pre-processing performed to construct a modest dataset to overcome the lack of annotated data. Also proposed end- to-end speech recognition model, the design of the model was formed using Convolution Neural Networks. The Sudanese dialect dataset would be a stepping stone to enable future Natural Language Processing research targeting the dialect. The designed model provided some insights into the current recognition task and reached an average Label Error Rate of 73.67%.

☆ Continual Contrastive Finetuning Improves Low-Resource Relation Extraction

Wenxuan Zhou, Sheng Zhang, Tristan Naumann, Muhao Chen, Hoifung Poon

Relation extraction (RE), which has relied on structurally annotated corpora for model training, has been particularly challenging in low-resource scenarios and domains. Recent literature has tackled low-resource RE by self-supervised learning, where the solution involves pretraining the relation embedding by RE-based objective and finetuning on labeled data by classification-based objective. However, a critical challenge to this approach is the gap in objectives, which prevents the RE model from fully utilizing the knowledge in pretrained representations. In this paper, we aim at bridging the gap and propose to pretrain and finetune the RE model using consistent objectives of contrastive learning. Since in this kind of representation learning paradigm, one relation may easily form multiple clusters in the representation space, we further propose a multi-center contrastive loss that allows one relation to form multiple clusters to better align with pretraining. Experiments on two document-level RE datasets, BioRED and Re-DocRED, demonstrate the effectiveness of our method. Particularly, when using 1% end-task training data, our method outperforms PLM-based RE classifier by 10.5% and 5.8% on the two datasets, respectively.

☆ Attend to the Right Context: A Plug-and-Play Module for Content-Controllable Summarization

Wen Xiao, Lesly Miculicich, Yang Liu, Pengcheng He, Giuseppe Carenini

Content-Controllable Summarization generates summaries focused on the given controlling signals. Due to the lack of large-scale training corpora for the task, we propose a plug-and-play module RelAttn to adapt any general summarizers to the content-controllable summarization task. RelAttn first identifies the relevant content in the source documents, and then makes the model attend to the right context by directly steering the attention weight. We further apply an unsupervised online adaptive parameter searching algorithm to determine the degree of control in the zero-shot setting, while such parameters are learned in the few-shot setting. By applying the module to three backbone summarization models, experiments show that our method effectively improves all the summarizers, and outperforms the prefix-based method and a widely used plug-and-play model in both zero- and few-shot settings. Tellingly, more benefit is observed in the scenarios when more control is needed.

☆ 4D ASR: Joint modeling of CTC, Attention, Transducer, and Mask-Predict decoders ICASSP 2023

Yui Sudo, Muhammad Shakeel, Brian Yan, Jiatong Shi, Shinji Watanabe

The network architecture of end-to-end (E2E) automatic speech recognition (ASR) can be classified into several models, including connectionist temporal classification (CTC), recurrent neural network transducer (RNN-T), attention mechanism, and non-autoregressive mask-predict models. Since each of these network architectures has pros and cons, a typical use case is to switch these separate models depending on the application requirement, resulting in the increased overhead of maintaining all models. Several methods for integrating two of these complementary models to mitigate the overhead issue have been proposed; however, if we integrate more models, we will further benefit from these complementary models and realize broader applications with a single system. This paper proposes four-decoder joint modeling (4D) of CTC, attention, RNN-T, and mask-predict, which has the following three advantages: 1) The four decoders are jointly trained so that they can be easily switched depending on the application scenarios. 2) Joint training may bring model regularization and improve the model robustness thanks to their complementary properties. 3) Novel one-pass joint decoding methods using CTC, attention, and RNN-T further improves the performance. The experimental results showed that the proposed model consistently reduced the WER.

comment: Submitted to ICASSP 2023

☆ ZEROTOP: Zero-Shot Task-Oriented Semantic Parsing using Large Language Models

Dheeraj Mekala, Jason Wolfe, Subhro Roy

We explore the use of large language models (LLMs) for zero-shot semantic parsing. Semantic parsing involves mapping natural language utterances to task-specific meaning representations. Language models are generally trained on the publicly available text and code and cannot be expected to directly generalize to domain-specific parsing tasks in a zero-shot setting. In this work, we propose ZEROTOP, a zero-shot task-oriented parsing method that decomposes a semantic parsing problem into a set of abstractive and extractive question-answering (QA) problems, enabling us to leverage the ability of LLMs to zero-shot answer reading comprehension questions. For each utterance, we prompt the LLM with questions corresponding to its top-level intent and a set of slots and use the LLM generations to construct the target meaning representation. We observe that current LLMs fail to detect unanswerable questions; and as a result, cannot handle questions corresponding to missing slots. To address this problem, we fine-tune a language model on public QA datasets using synthetic negative samples. Experimental results show that our QA-based decomposition paired with the fine-tuned LLM can correctly parse ~16% of utterances in the MTOP dataset without requiring any annotated data.

☆ Reconstruction Probing

Najoung Kim, Jatin Khilnani, Alex Warstadt, Abed Qaddoumi

We propose reconstruction probing, a new analysis method for contextualized representations based on reconstruction probabilities in masked language models (MLMs). This method relies on comparing the reconstruction probabilities of tokens in a given sequence when conditioned on the representation of a single token that has been fully contextualized and when conditioned on only the decontextualized lexical prior of the model. This comparison can be understood as quantifying the contribution of contextualization towards reconstruction -- the difference in the reconstruction probabilities can only be attributed to the representational change of the single token induced by contextualization. We apply this analysis to three MLMs and find that contextualization boosts reconstructability of tokens that are close to the token being reconstructed in terms of linear and syntactic distance. Furthermore, we extend our analysis to finer-grained decomposition of contextualized representations, and we find that these boosts are largely attributable to static and positional embeddings at the input layer.

comment: Preprint

☆ OpineSum: Entailment-based self-training for abstractive opinion summarization

Annie Louis, Joshua Maynez

A typical product or place often has hundreds of reviews, and summarization of these texts is an important and challenging problem. Recent progress on abstractive summarization in domains such as news has been driven by supervised systems trained on hundreds of thousands of news articles paired with human-written summaries. However for opinion texts, such large scale datasets are rarely available. Unsupervised methods, self-training, and few-shot learning approaches bridge that gap. In this work, we present a novel self-training approach, OpineSum, for abstractive opinion summarization. The summaries in this approach are built using a novel application of textual entailment and capture the consensus of opinions across the various reviews for an item. This method can be used to obtain silver-standard summaries on a large scale and train both unsupervised and few-shot abstractive summarization systems. OpineSum achieves state-of-the-art performance in both settings.

☆ Multi-modal Molecule Structure-text Model for Text-based Retrieval and Editing

Shengchao Liu, Weili Nie, Chengpeng Wang, Jiarui Lu, Zhuoran Qiao, Ling Liu, Jian Tang, Chaowei Xiao, Anima Anandkumar

There is increasing adoption of artificial intelligence in drug discovery. However, existing works use machine learning to mainly utilize the chemical structures of molecules yet ignore the vast textual knowledge available in chemistry. Incorporating textual knowledge enables us to realize new drug design objectives, adapt to text-based instructions, and predict complex biological activities. We present a multi-modal molecule structure-text model, MoleculeSTM, by jointly learning molecule's chemical structures and textual descriptions via a contrastive learning strategy. To train MoleculeSTM, we construct the largest multi-modal dataset to date, namely PubChemSTM, with over 280K chemical structure-text pairs. To demonstrate the effectiveness and utility of MoleculeSTM, we design two challenging zero-shot tasks based on text instructions, including structure-text retrieval and molecule editing. MoleculeSTM possesses two main properties: open vocabulary and compositionality via natural language. In experiments, MoleculeSTM obtains the state-of-the-art generalization ability to novel biochemical concepts across various benchmarks.

☆ Multi-hop Evidence Retrieval for Cross-document Relation Extraction

Keming Lu, I-Hung Hsu, Wenxuan Zhou, Mingyu Derek Ma, Muhao Chen

Relation Extraction (RE) has been extended to cross-document scenarios because many relations are not simply described in a single document. This inevitably brings the challenge of efficient open-space evidence retrieval to support the inference of cross-document relations, along with the challenge of multi-hop reasoning on top of entities and evidence scattered in an open set of documents. To combat these challenges, we propose Mr.CoD, a multi-hop evidence retrieval method based on evidence path mining and ranking with adapted dense retrievers. We explore multiple variants of retrievers to show evidence retrieval is an essential part in cross-document RE. Experiments on CodRED show that evidence retrieval with Mr.Cod effectively acquires cross-document evidence that essentially supports open-setting cross-document RE. Additionally, we show that Mr.CoD facilitates evidence retrieval and boosts end-to-end RE performance with effective multi-hop reasoning in both closed and open settings of RE.

comment: Work in progress

☆ SERENGETI: Massively Multilingual Language Models for Africa

Ife Adebara, AbdelRahim Elmadany, Muhammad Abdul-Mageed, Alcides Alcoba Inciarte

Multilingual language models (MLMs) acquire valuable, generalizable linguistic information during pretraining and have advanced the state of the art on task-specific finetuning. So far, only ~ 28 out of ~2,000 African languages are covered in existing language models. We ameliorate this limitation by developing SERENGETI, a set of massively multilingual language model that covers 517 African languages and language varieties. We evaluate our novel models on eight natural language understanding tasks across 20 datasets, comparing to four MLMs that each cover any number of African languages. SERENGETI outperforms other models on 11 datasets across the eights tasks and achieves 82.27 average F-1. We also perform error analysis on our models' performance and show the influence of mutual intelligibility when the models are applied under zero-shot settings. We will publicly release our models for research.

comment: Work in Progress

☆ Can NLI Provide Proper Indirect Supervision for Low-resource Biomedical Relation Extraction?

Jiashu Xu, Mingyu Derek Ma, Muhao Chen

Two key obstacles in biomedical relation extraction (RE) are the scarcity of annotations and the prevalence of instances without explicitly pre-defined labels due to low annotation coverage. Existing approaches, which treat biomedical RE as a multi-class classification task, often result in poor generalization in low-resource settings and do not have the ability to make selective prediction on unknown cases but give a guess from seen relations, hindering the applicability of those approaches. We present NBR, which converts biomedical RE as natural language inference formulation through indirect supervision. By converting relations to natural language hypotheses, NBR is capable of exploiting semantic cues to alleviate annotation scarcity. By incorporating a ranking-based loss that implicitly calibrates abstinent instances, NBR learns a clearer decision boundary and is instructed to abstain on uncertain instances. Extensive experiments on three widely-used biomedical RE benchmarks, namely ChemProt, DDI and GAD, verify the effectiveness of NBR in both full-set and low-resource regimes. Our analysis demonstrates that indirect supervision benefits biomedical RE even when a domain gap exists, and combining NLI knowledge with biomedical knowledge leads to the best performance gains.

comment: 16 pages

☆ MultiInstruct: Improving Multi-Modal Zero-Shot Learning via Instruction Tuning

Zhiyang Xu, Ying Shen, Lifu Huang

Instruction tuning, a new learning paradigm that fine-tunes pre-trained language models on tasks specified through instructions, has shown promising zero-shot performance on various natural language processing tasks. However, it's still not explored for vision and multimodal tasks. In this work, we introduce MultiInstruct, the first multimodal instruction tuning benchmark dataset that consists of 47 diverse multimodal tasks covering 11 broad categories. Each task is designed at least with 5,000 instances (input-out pairs) from existing open-source datasets and 5 expert-written instructions. We take OFA as the base pre-trained model for multimodal instruction tuning, and to improve its performance, we explore multiple transfer learning strategies to leverage the large-scale Natural Instructions dataset. Experimental results demonstrate its strong zero-shot performance on various unseen multimodal tasks and the benefit of transfer learning from text-only instructions. We also design a new evaluation metric: Sensitivity, to evaluate how sensitive the model is to the variety of instructions. Our results indicate that the model is less sensitive to the varying instructions after finetuning on a diverse set of tasks and instructions for each task.

comment: 16 pages, 10 tables, 4 figures

☆ ImPaKT: A Dataset for Open-Schema Knowledge Base Construction

Luke Vilnis, Zach Fisher, Bhargav Kanagal, Patrick Murray, Sumit Sanghai

Large language models have ushered in a golden age of semantic parsing. The seq2seq paradigm allows for open-schema and abstractive attribute and relation extraction given only small amounts of finetuning data. Language model pretraining has simultaneously enabled great strides in natural language inference, reasoning about entailment and implication in free text. These advances motivate us to construct ImPaKT, a dataset for open-schema information extraction, consisting of around 2500 text snippets from the C4 corpus, in the shopping domain (product buying guides), professionally annotated with extracted attributes, types, attribute summaries (attribute schema discovery from idiosyncratic text), many-to-one relations between compound and atomic attributes, and implication relations. We release this data in hope that it will be useful in fine tuning semantic parsers for information extraction and knowledge base construction across a variety of domains. We evaluate the power of this approach by fine-tuning the open source UL2 language model on a subset of the dataset, extracting a set of implication relations from a corpus of product buying guides, and conducting human evaluations of the resulting predictions.

comment: 14 pages. Preprint

☆ Uncontrolled Lexical Exposure Leads to Overestimation of Compositional Generalization in Pretrained Models

Najoung Kim, Tal Linzen, Paul Smolensky

Human linguistic capacity is often characterized by compositionality and the generalization it enables -- human learners can produce and comprehend novel complex expressions by composing known parts. Several benchmarks exploit distributional control across training and test to gauge compositional generalization, where certain lexical items only occur in limited contexts during training. While recent work using these benchmarks suggests that pretrained models achieve impressive generalization performance, we argue that exposure to pretraining data may break the aforementioned distributional control. Using the COGS benchmark of Kim and Linzen (2020), we test two modified evaluation setups that control for this issue: (1) substituting context-controlled lexical items with novel character sequences, and (2) substituting them with special tokens represented by novel embeddings. We find that both of these setups lead to lower generalization performance in T5 (Raffel et al., 2020), suggesting that previously reported results have been overestimated due to uncontrolled lexical exposure during pretraining. The performance degradation is more extreme with novel embeddings, and the degradation increases with the amount of pretraining data, highlighting an interesting case of inverse scaling.

comment: Preprint

☆ How Does Beam Search improve Span-Level Confidence Estimation in Generative Sequence Labeling?

Kazuma Hashimoto, Iftekhar Naim, Karthik Raman

Text-to-text generation models have increasingly become the go-to solution for a wide variety of sequence labeling tasks (e.g., entity extraction and dialog slot filling). While most research has focused on the labeling accuracy, a key aspect -- of vital practical importance -- has slipped through the cracks: understanding model confidence. More specifically, we lack a principled understanding of how to reliably gauge the confidence of a model in its predictions for each labeled span. This paper aims to provide some empirical insights on estimating model confidence for generative sequence labeling. Most notably, we find that simply using the decoder's output probabilities is not the best in realizing well-calibrated confidence estimates. As verified over six public datasets of different tasks, we show that our proposed approach -- which leverages statistics from top-$k$ predictions by a beam search -- significantly reduces calibration errors of the predictions of a generative sequence labeling model.

☆ Learning List-Level Domain-Invariant Representations for Ranking

Ruicheng Xian, Honglei Zhuang, Zhen Qin, Hamed Zamani, Jing Lu, Ji Ma, Kai Hui, Han Zhao, Xuanhui Wang, Michael Bendersky

Domain adaptation aims to transfer the knowledge acquired by models trained on (data-rich) source domains to (low-resource) target domains, for which a popular method is invariant representation learning. While they have been studied extensively for classification and regression problems, how they apply to ranking problems, where the data and metrics have a list structure, is not well understood. Theoretically, we establish a domain adaptation generalization bound for ranking under listwise metrics such as MRR and NDCG. The bound suggests an adaptation method via learning list-level domain-invariant feature representations, whose benefits are empirically demonstrated by unsupervised domain adaptation experiments on real-world ranking tasks, including passage reranking. A key message is that for domain adaptation, the representations should be analyzed at the same level at which the metric is computed, as we show that learning invariant representations at the list level is most effective for adaptation on ranking problems.

☆ ORCA: A Challenging Benchmark for Arabic Language Understanding

AbdelRahim Elmadany, El Moatez Billah Nagoudi, Muhammad Abdul-Mageed

Due to their crucial role in all NLP, several benchmarks have been proposed to evaluate pretrained language models. In spite of these efforts, no public benchmark of diverse nature currently exists for evaluation of Arabic. This makes it challenging to measure progress for both Arabic and multilingual language models. This challenge is compounded by the fact that any benchmark targeting Arabic needs to take into account the fact that Arabic is not a single language but rather a collection of languages and varieties. In this work, we introduce ORCA, a publicly available benchmark for Arabic language understanding evaluation. ORCA is carefully constructed to cover diverse Arabic varieties and a wide range of challenging Arabic understanding tasks exploiting 60 different datasets across seven NLU task clusters. To measure current progress in Arabic NLU, we use ORCA to offer a comprehensive comparison between 18 multilingual and Arabic language models. We also provide a public leaderboard with a unified single-number evaluation metric (ORCA score) to facilitate future research.

comment: All authors contributed equally

☆ JASMINE: Arabic GPT Models for Few-Shot Learning

El Moatez Billah Nagoudi, Muhammad Abdul-Mageed, AbdelRahim Elmadany, Alcides Alcoba Inciarte, Md Tawkat Islam Khondaker

Task agnostic generative pretraining (GPT) has recently proved promising for zero- and few-shot learning, gradually diverting attention from the expensive supervised learning paradigm. Although the community is accumulating knowledge as to capabilities of English-language autoregressive models such as GPT-3 adopting this generative approach, scholarship about these models remains acutely Anglocentric. Consequently, the community currently has serious gaps in its understanding of this class of models, their potential, and their societal impacts in diverse settings, linguistic traditions, and cultures. To alleviate this issue for Arabic, a collection of diverse languages and language varieties with more than $400$ million population, we introduce JASMINE, a suite of powerful Arabic autoregressive Transformer language models ranging in size between 300 million-13 billion parameters. We pretrain our new models with large amounts of diverse data (400GB of text) from different Arabic varieties and domains. We evaluate JASMINE extensively in both intrinsic and extrinsic settings, using a comprehensive benchmark for zero- and few-shot learning across a wide range of NLP tasks. We also carefully develop and release a novel benchmark for both automated and human evaluation of Arabic autoregressive models focused at investigating potential social biases, harms, and toxicity in these models. We aim to responsibly release our models with interested researchers, along with code for experimenting with them

☆ CORRPUS: Detecting Story Inconsistencies via Codex-Bootstrapped Neurosymbolic Reasoning

Yijiang River Dong, Lara J. Martin, Chris Callison-Burch

Story generation and understanding -- as with all NLG/NLU tasks -- has seen a surge in neurosymbolic work. Researchers have recognized that, while large language models (LLMs) have tremendous utility, they can be augmented with symbolic means to be even better and to make up for any flaws that the neural networks might have. However, symbolic methods are extremely costly in terms of the amount of time and expertise needed to create them. In this work, we capitalize on state-of-the-art Code-LLMs, such as Codex, to bootstrap the use of symbolic methods for tracking the state of stories and aiding in story understanding. We show that our CoRRPUS system and abstracted prompting procedures can beat current state-of-the-art structured LLM techniques on pre-existing story understanding tasks (bAbI task 2 and Re^3) with minimal hand engineering. We hope that this work can help highlight the importance of symbolic representations and specialized prompting for LLMs as these models require some guidance for performing reasoning tasks properly.

☆ PropSegmEnt: A Large-Scale Corpus for Proposition-Level Segmentation and Entailment Recognition

Sihao Chen, Senaka Buthpitiya, Alex Fabrikant, Dan Roth, Tal Schuster

The widely studied task of Natural Language Inference (NLI) requires a system to recognize whether one piece of text is textually entailed by another, i.e. whether the entirety of its meaning can be inferred from the other. In current NLI datasets and models, textual entailment relations are typically defined on the sentence- or paragraph-level. However, even a simple sentence often contains multiple propositions, i.e. distinct units of meaning conveyed by the sentence. As these propositions can carry different truth values in the context of a given premise, we argue for the need to recognize the textual entailment relation of each proposition in a sentence individually. We propose PropSegmEnt, a corpus of over 35K propositions annotated by expert human raters. Our dataset structure resembles the tasks of (1) segmenting sentences within a document to the set of propositions, and (2) classifying the entailment relation of each proposition with respect to a different yet topically-aligned document, i.e. documents describing the same event or entity. We establish strong baselines for the segmentation and entailment tasks. Through case studies on summary hallucination detection and document-level NLI, we demonstrate that our conceptual framework is potentially useful for understanding and explaining the compositionality of NLI labels.

☆ ToL: A Tensor of List-Based Unified Computation Model

Hongxiao Li, Wanling Gao, Lei Wang, Jianfeng Zhan

Previous computation models either have equivalent abilities in representing all computations but fail to provide primitive operators for programming complex algorithms or lack generalized expression ability to represent newly-added computations. This article presents a unified computation model with generalized expression ability and a concise set of primitive operators for programming high-level algorithms. We propose a unified data abstraction -- Tensor of List, and offer a unified computation model based on Tensor of List, which we call the ToL model (in short, ToL). ToL introduces five atomic computations that can represent any elementary computation by finite composition, ensured with strict formal proof. Based on ToL, we design a pure-functional language -- ToLang. ToLang provides a concise set of primitive operators that can be used to program complex big data and AI algorithms. Our evaluations show ToL has generalized expression ability and a built-in performance indicator, born with a strictly defined computation metric -- elementary operation count (EOPs), consistent with FLOPs within a small error range.

☆ Spoken Language Understanding for Conversational AI: Recent Advances and Future Direction

Soyeon Caren Han, Siqu Long, Henry Weld, Josiah Poon

When a human communicates with a machine using natural language on the web and online, how can it understand the human's intention and semantic context of their talk? This is an important AI task as it enables the machine to construct a sensible answer or perform a useful action for the human. Meaning is represented at the sentence level, identification of which is known as intent detection, and at the word level, a labelling task called slot filling. This dual-level joint task requires innovative thinking about natural language and deep learning network design, and as a result, many approaches and models have been proposed and applied. This tutorial will discuss how the joint task is set up and introduce Spoken Language Understanding/Natural Language Understanding (SLU/NLU) with Deep Learning techniques. We will cover the datasets, experiments and metrics used in the field. We will describe how the machine uses the latest NLP and Deep Learning techniques to address the joint task, including recurrent and attention-based Transformer networks and pre-trained models (e.g. BERT). We will then look in detail at a network that allows the two levels of the task, intent classification and slot filling, to interact to boost performance explicitly. We will do a code demonstration of a Python notebook for this model and attendees will have an opportunity to watch coding demo tasks on this joint NLU to further their understanding.

comment: Accepted by TheWebConf2023. arXiv admin note: substantial text overlap with arXiv:2101.08091

☆ Beyond Contrastive Learning: A Variational Generative Model for Multilingual Retrieval

John Wieting, Jonathan H. Clark, William W. Cohen, Graham Neubig, Taylor Berg-Kirkpatrick

Contrastive learning has been successfully used for retrieval of semantically aligned sentences, but it often requires large batch sizes or careful engineering to work well. In this paper, we instead propose a generative model for learning multilingual text embeddings which can be used to retrieve or score sentence pairs. Our model operates on parallel data in $N$ languages and, through an approximation we introduce, efficiently encourages source separation in this multilingual setting, separating semantic information that is shared between translations from stylistic or language-specific variation. We show careful large-scale comparisons between contrastive and generation-based approaches for learning multilingual text embeddings, a comparison that has not been done to the best of our knowledge despite the popularity of these approaches. We evaluate this method on a suite of tasks including semantic similarity, bitext mining, and cross-lingual question retrieval -- the last of which we introduce in this paper. Overall, our Variational Multilingual Source-Separation Transformer (VMSST) model outperforms both a strong contrastive and generative baseline on these tasks.

☆ Tracing and Removing Data Errors in Natural Language Generation Datasets

Faisal Ladhak, Esin Durmus, Tatsunori Hashimoto

Recent work has identified noisy and misannotated data as a core cause of hallucinations and unfaithful outputs in Natural Language Generation (NLG) tasks. Consequently, identifying and removing these examples is a key open challenge in creating reliable NLG systems. In this work, we introduce a framework to identify and remove low-quality training instances that lead to undesirable outputs, such as faithfulness errors in text summarization. We show that existing approaches for error tracing, such as gradient-based influence measures, do not perform reliably for detecting faithfulness errors in summarization. We overcome the drawbacks of existing error tracing methods through a new, contrast-based estimate that compares undesired generations to human-corrected outputs. Our proposed method can achieve a mean average precision of 0.91 across synthetic tasks with known ground truth and can achieve a two-fold reduction in hallucinations on a real entity hallucination evaluation on the NYT dataset.

☆ MoralDial: A Framework to Train and Evaluate Moral Dialogue Systems via Constructing Moral Discussions

Hao Sun, Zhexin Zhang, Fei Mi, Yasheng Wang, Wei Liu, Jianwei Cui, Bin Wang, Qun Liu, Minlie Huang

Morality in dialogue systems has raised great attention in research recently. A moral dialogue system could better connect users and enhance conversation engagement by gaining users' trust. In this paper, we propose a framework, MoralDial to train and evaluate moral dialogue systems. In our framework, we first explore the communication mechanisms of morality and resolve expressed morality into four sub-modules. The sub-modules indicate the roadmap for building a moral dialogue system. Based on that, we design a simple yet effective method: constructing moral discussions from Rules of Thumb (RoTs) between simulated specific users and the dialogue system. The constructed discussion consists of expressing, explaining, and revising the moral views in dialogue exchanges, which makes conversational models learn morality well in a natural manner. Furthermore, we propose a novel evaluation method in the framework. We evaluate the multiple aspects of morality by judging the relation between dialogue responses and RoTs in discussions, where the multifaceted nature of morality is particularly considered. Automatic and manual experiments demonstrate that our framework is promising to train and evaluate moral dialogue systems.

☆ Integrating Heterogeneous Domain Information into Relation Extraction: A Case Study on Drug-Drug Interaction Extraction

Masaki Asada

The development of deep neural networks has improved representation learning in various domains, including textual, graph structural, and relational triple representations. This development opened the door to new relation extraction beyond the traditional text-oriented relation extraction. However, research on the effectiveness of considering multiple heterogeneous domain information simultaneously is still under exploration, and if a model can take an advantage of integrating heterogeneous information, it is expected to exhibit a significant contribution to many problems in the world. This thesis works on Drug-Drug Interactions (DDIs) from the literature as a case study and realizes relation extraction utilizing heterogeneous domain information. First, a deep neural relation extraction model is prepared and its attention mechanism is analyzed. Next, a method to combine the drug molecular structure information and drug description information to the input sentence information is proposed, and the effectiveness of utilizing drug molecular structures and drug descriptions for the relation extraction task is shown. Then, in order to further exploit the heterogeneous information, drug-related items, such as protein entries, medical terms and pathways are collected from multiple existing databases and a new data set in the form of a knowledge graph (KG) is constructed. A link prediction task on the constructed data set is conducted to obtain embedding representations of drugs that contain the heterogeneous domain information. Finally, a method that integrates the input sentence information and the heterogeneous KG information is proposed. The proposed model is trained and evaluated on a widely used data set, and as a result, it is shown that utilizing heterogeneous domain information significantly improves the performance of relation extraction from the literature.

comment: PhD Thesis

☆ Zero-shot Triplet Extraction by Template Infilling

Bosung Kim, Hayate Iso, Nikita Bhutani, Estevam Hruschka, Ndapa Nakashole

Triplet extraction aims to extract entities and their corresponding relations in unstructured text. Most existing methods train an extraction model on high-quality training data, and hence are incapable of extracting relations that were not observed during training. Generalizing the model to unseen relations typically requires fine-tuning on synthetic training data which is often noisy and unreliable. In this paper, we argue that reducing triplet extraction to a template filling task over a pre-trained language model can equip the model with zero-shot learning capabilities and enable it to leverage the implicit knowledge in the language model. Embodying these ideas, we propose a novel framework, ZETT (ZEro-shot Triplet extraction by Template infilling), that is based on end-to-end generative transformers. Our experiments show that without any data augmentation or pipeline systems, ZETT can outperform previous state-of-the-art models with 25% less parameters. We further show that ZETT is more robust in detecting entities and can be incorporated with automatically generated templates for relations.

comment: 12 pages, 2 figures

☆ Extractive Text Summarization Using Generalized Additive Models with Interactions for Sentence Selection

Vinícius Camargo da Silva, João Paulo Papa, Kelton Augusto Pontara da Costa

Automatic Text Summarization (ATS) is becoming relevant with the growth of textual data; however, with the popularization of public large-scale datasets, some recent machine learning approaches have focused on dense models and architectures that, despite producing notable results, usually turn out in models difficult to interpret. Given the challenge behind interpretable learning-based text summarization and the importance it may have for evolving the current state of the ATS field, this work studies the application of two modern Generalized Additive Models with interactions, namely Explainable Boosting Machine and GAMI-Net, to the extractive summarization problem based on linguistic features and binary classification.

☆ Analyzing Semantic Faithfulness of Language Models via Input Intervention on Conversational Question Answering

Akshay Chaturvedi, Swarnadeep Bhar, Soumadeep Saha, Utpal Garain, Nicholas Asher

Transformer-based language models have been shown to be highly effective for several NLP tasks. In this paper, we consider three transformer models, BERT, RoBERTa, and XLNet, in both small and large version, and investigate how faithful their representations are with respect to the semantic content of texts. We formalize a notion of semantic faithfulness, in which the semantic content of a text should causally figure in a model's inferences in question answering. We then test this notion by observing a model's behavior on answering questions about a story after performing two novel semantic interventions -- deletion intervention and negation intervention. While transformer models achieve high performance on standard question answering tasks, we show that they fail to be semantically faithful once we perform these interventions for a significant number of cases (~50% for deletion intervention, and ~20% drop in accuracy for negation intervention). We then propose an intervention-based training regime that can mitigate the undesirable effects for deletion intervention by a significant margin (from ~50% to ~6%). We analyze the inner-workings of the models to better understand the effectiveness of intervention-based training for deletion intervention. But we show that this training does not attenuate other aspects of semantic unfaithfulness such as the models' inability to deal with negation intervention or to capture the predicate-argument structure of texts. We also test InstructGPT, via prompting, for its ability to handle the two interventions and to capture predicate-argument structure. While InstructGPT models do achieve very high performance on predicate-argument structure task, they fail to respond adequately to our deletion and negation interventions.

comment: 27 pages, 4 figures

☆ Automatic Emotion Modelling in Written Stories

Lukas Christ, Shahin Amiriparian, Manuel Milling, Ilhan Aslan, Björn W. Schuller

Telling stories is an integral part of human communication which can evoke emotions and influence the affective states of the audience. Automatically modelling emotional trajectories in stories has thus attracted considerable scholarly interest. However, as most existing works have been limited to unsupervised dictionary-based approaches, there is no labelled benchmark for this task. We address this gap by introducing continuous valence and arousal annotations for an existing dataset of children's stories annotated with discrete emotion categories. We collect additional annotations for this data and map the originally categorical labels to the valence and arousal space. Leveraging recent advances in Natural Language Processing, we propose a set of novel Transformer-based methods for predicting valence and arousal signals over the course of written stories. We explore several strategies for fine-tuning a pretrained ELECTRA model and study the benefits of considering a sentence's context when inferring its emotionality. Moreover, we experiment with additional LSTM and Transformer layers. The best configuration achieves a Concordance Correlation Coefficient (CCC) of .7338 for valence and .6302 for arousal on the test set, demonstrating the suitability of our proposed approach. Our code and additional annotations are made available at https://github.com/lc0197/emotion_modelling_stories.

comment: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

☆ Contrastive Distillation Is a Sample-Efficient Self-Supervised Loss Policy for Transfer Learning

Chris Lengerich, Gabriel Synnaeve, Amy Zhang, Hugh Leather, Kurt Shuster, François Charton, Charysse Redwood

Traditional approaches to RL have focused on learning decision policies directly from episodic decisions, while slowly and implicitly learning the semantics of compositional representations needed for generalization. While some approaches have been adopted to refine representations via auxiliary self-supervised losses while simultaneously learning decision policies, learning compositional representations from hand-designed and context-independent self-supervised losses (multi-view) still adapts relatively slowly to the real world, which contains many non-IID subspaces requiring rapid distribution shift in both time and spatial attention patterns at varying levels of abstraction. In contrast, supervised language model cascades have shown the flexibility to adapt to many diverse manifolds, and hints of self-learning needed for autonomous task transfer. However, to date, transfer methods for language models like few-shot learning and fine-tuning still require human supervision and transfer learning using self-learning methods has been underexplored. We propose a self-supervised loss policy called contrastive distillation which manifests latent variables with high mutual information with both source and target tasks from weights to tokens. We show how this outperforms common methods of transfer learning and suggests a useful design axis of trading off compute for generalizability for online transfer. Contrastive distillation is improved through sampling from memory and suggests a simple algorithm for more efficiently sampling negative examples for contrastive losses than random sampling.

☆ What do LLMs Know about Financial Markets? A Case Study on Reddit Market Sentiment Analysis

Xiang Deng, Vasilisa Bashlovkina, Feng Han, Simon Baumgartner, Michael Bendersky

Market sentiment analysis on social media content requires knowledge of both financial markets and social media jargon, which makes it a challenging task for human raters. The resulting lack of high-quality labeled data stands in the way of conventional supervised learning methods. Instead, we approach this problem using semi-supervised learning with a large language model (LLM). Our pipeline generates weak financial sentiment labels for Reddit posts with an LLM and then uses that data to train a small model that can be served in production. We find that prompting the LLM to produce Chain-of-Thought summaries and forcing it through several reasoning paths helps generate more stable and accurate labels, while using a regression loss further improves distillation quality. With only a handful of prompts, the final model performs on par with existing supervised models. Though production applications of our model are limited by ethical considerations, the model's competitive performance points to the great potential of using LLMs for tasks that otherwise require skill-intensive annotation.

☆ Language models are better than humans at next-token prediction

Buck Shlegeris, Fabien Roger, Lawrence Chan, Euan McLean

Current language models are considered to have sub-human capabilities at natural language tasks like question-answering or writing code. However, language models are not trained to perform well at these tasks, they are trained to accurately predict the next token given previous tokes in tokenized text. It is not clear whether language models are better or worse than humans at next token prediction. To try to answer this question, we performed two distinct experiments to directly compare humans and language models on this front: one measuring top-1 accuracy and the other measuring perplexity. In both experiments, we find humans to be consistently \emph{worse} than even relatively small language models like GPT3-Ada at next-token prediction.

☆ KL Regularized Normalization Framework for Low Resource Tasks

Neeraj Kumar, Ankur Narang, Brejesh Lall

Large pre-trained models, such as Bert, GPT, and Wav2Vec, have demonstrated great potential for learning representations that are transferable to a wide variety of downstream tasks . It is difficult to obtain a large quantity of supervised data due to the limited availability of resources and time. In light of this, a significant amount of research has been conducted in the area of adopting large pre-trained datasets for diverse downstream tasks via fine tuning, linear probing, or prompt tuning in low resource settings. Normalization techniques are essential for accelerating training and improving the generalization of deep neural networks and have been successfully used in a wide variety of applications. A lot of normalization techniques have been proposed but the success of normalization in low resource downstream NLP and speech tasks is limited. One of the reasons is the inability to capture expressiveness by rescaling parameters of normalization. We propose KullbackLeibler(KL) Regularized normalization (KL-Norm) which make the normalized data well behaved and helps in better generalization as it reduces over-fitting, generalises well on out of domain distributions and removes irrelevant biases and features with negligible increase in model parameters and memory overheads. Detailed experimental evaluation on multiple low resource NLP and speech tasks, demonstrates the superior performance of KL-Norm as compared to other popular normalization and regularization techniques.

comment: arXiv admin note: text overlap with arXiv:2106.05469 by other authors

☆ A Mutation-based Text Generation for Adversarial Machine Learning Applications

Jesus Guerrero, Gongbo Liang, Izzat Alsmadi

Many natural language related applications involve text generation, created by humans or machines. While in many of those applications machines support humans, yet in few others, (e.g. adversarial machine learning, social bots and trolls) machines try to impersonate humans. In this scope, we proposed and evaluated several mutation-based text generation approaches. Unlike machine-based generated text, mutation-based generated text needs human text samples as inputs. We showed examples of mutation operators but this work can be extended in many aspects such as proposing new text-based mutation operators based on the nature of the application.

♻ ☆ GROOT: Corrective Reward Optimization for Generative Sequential Labeling

Kazuma Hashimoto, Karthik Raman

Sequential labeling is a fundamental NLP task, forming the backbone of many applications. Supervised learning of Seq2Seq models has shown great success on these problems. However, the training objectives are still significantly disconnected with the metrics and desiderata we care about in practice. For example, a practical sequence tagging application may want to optimize for a certain precision-recall trade-off (of the top-k predictions) which is quite different from the standard objective of maximizing the likelihood of the gold labeled sequence. Thus to bridge this gap, we propose GROOT -- a simple yet effective framework for Generative Reward Optimization Of Text sequences. GROOT works by training a generative sequential labeling model to match the decoder output distribution with that of the (black-box) reward function. Using an iterative training regime, we first generate prediction candidates, then correct errors in them, and finally contrast those candidates (based on their reward values). As demonstrated via extensive experiments on four public benchmarks, GROOT significantly improves all reward metrics. Furthermore, GROOT leads to improvements of the overall decoder distribution as evidenced by the quality gains of the top-$k$ candidates.

♻ ☆ Does human speech follow Benford's Law?

Leo Hsu, Visar Berisha

Researchers have observed that the frequencies of leading digits in many man-made and naturally occurring datasets follow a logarithmic curve, with digits that start with the number 1 accounting for $\sim 30\%$ of all numbers in the dataset and digits that start with the number 9 accounting for $\sim 5\%$ of all numbers in the dataset. This phenomenon, known as Benford's Law, is highly repeatable and appears in lists of numbers from electricity bills, stock prices, tax returns, house prices, death rates, lengths of rivers, and naturally occurring images. In this paper we demonstrate that human speech spectra also follow Benford's Law on average. That is, when averaged over many speakers, the frequencies of leading digits in speech magnitude spectra follow this distribution, although with some variability at the individual sample level. We use this observation to motivate a new set of features that can be efficiently extracted from speech and demonstrate that these features can be used to classify between human speech and synthetic speech.

♻ ☆ Multimodal Hate Speech Detection from Bengali Memes and Texts

Md. Rezaul Karim, Sumon Kanti Dey, Tanhim Islam, Md. Shajalal, Bharathi Raja Chakravarthi

Numerous machine learning (ML) and deep learning (DL)-based approaches have been proposed to utilize textual data from social media for anti-social behavior analysis like cyberbullying, fake news detection, and identification of hate speech mainly for highly-resourced languages such as English. However, despite having a lot of diversity and millions of native speakers, some languages like Bengali are under-resourced, which is due to a lack of computational resources for natural language processing (NLP). Similar to other languages, Bengali social media contents also include images along with texts (e.g., multimodal memes are posted by embedding short texts into images on Facebook). Therefore, only the textual data is not enough to judge them since images might give extra context to make a proper judgement. This paper is about hate speech detection from multimodal Bengali memes and texts. We prepared the only multimodal hate speech dataset for-a-kind of problem for Bengali, which we use to train state-of-the-art neural architectures (e.g., Bi-LSTM/Conv-LSTM with word embeddings, ConvNets + pre-trained language models, e.g., monolingual Bangla BERT, multilingual BERT-cased/uncased, and XLM-RoBERTa) to jointly analyze textual and visual information for hate speech detection. Conv-LSTM and XLM-RoBERTa models performed best for texts, yielding F1 scores of 0.78 and 0.82, respectively. As of memes, ResNet-152 and DenseNet-161 models yield F1 scores of 0.78 and 0.79, respectively. As for multimodal fusion, XLM-RoBERTa + DenseNet-161 performed the best, yielding an F1 score of 0.83. Our study suggests that text modality is most useful for hate speech detection, while memes are moderately useful.

comment: arXiv admin note: text overlap with arXiv:2107.00648 by other authors

♻ ☆ Towards Linguistically Informed Multi-Objective Pre-Training for Natural Language Inference

Maren Pielka, Svetlana Schmidt, Lisa Pucknat, Rafet Sifa

We introduce a linguistically enhanced combination of pre-training methods for transformers. The pre-training objectives include POS-tagging, synset prediction based on semantic knowledge graphs, and parent prediction based on dependency parse trees. Our approach achieves competitive results on the Natural Language Inference task, compared to the state of the art. Specifically for smaller models, the method results in a significant performance boost, emphasizing the fact that intelligent pre-training can make up for fewer parameters and help building more efficient models. Combining POS-tagging and synset prediction yields the overall best results.

♻ ☆ Is Your Model Sensitive? SPeDaC: A New Benchmark for Detecting and Classifying Sensitive Personal Data

Gaia Gambarelli, Aldo Gangemi, Rocco Tripodi

In recent years, there has been an exponential growth of applications, including dialogue systems, that handle sensitive personal information. This has brought to light the extremely important issue of personal data protection in virtual environments. Sensitive Information Detection (SID) approaches different domains and languages in literature. However, if we refer to the personal data domain, a shared benchmark or the absence of an available labeled resource makes comparison with the state-of-the-art difficult. We introduce and release SPeDaC , a new annotated resource for the identification of sensitive personal data categories in the English language. SPeDaC enables the evaluation of computational models for three different SID subtasks with increasing levels of complexity. SPeDaC 1 regards binary classification, a model has to detect if a sentence contains sensitive information or not; whereas, in SPeDaC 2 we collected labeled sentences using 5 categories that relate to macro-domains of personal information; in SPeDaC 3, the labeling is fine-grained (61 personal data categories). We conduct an extensive evaluation of the resource using different state-of-the-art-classifiers. The results show that SPeDaC is challenging, particularly with regard to fine-grained classification. The transformer models achieve the best results (acc. RoBERTa on SPeDaC 1 = 98.20%, DeBERTa on SPeDaC 2 = 95.81% and SPeDaC 3 = 77.63%).

comment: 25 pages, 4 figures, 12 tables

♻ ☆ In conversation with Artificial Intelligence: aligning language models with human values

Atoosa Kasirzadeh, Iason Gabriel

Large-scale language technologies are increasingly used in various forms of communication with humans across different contexts. One particular use case for these technologies is conversational agents, which output natural language text in response to prompts and queries. This mode of engagement raises a number of social and ethical questions. For example, what does it mean to align conversational agents with human norms or values? Which norms or values should they be aligned with? And how can this be accomplished? In this paper, we propose a number of steps that help answer these questions. We start by developing a philosophical analysis of the building blocks of linguistic communication between conversational agents and human interlocutors. We then use this analysis to identify and formulate ideal norms of conversation that can govern successful linguistic communication between humans and conversational agents. Furthermore, we explore how these norms can be used to align conversational agents with human values across a range of different discursive domains. We conclude by discussing the practical implications of our proposal for the design of conversational agents that are aligned with these norms and values.

comment: Accepted for publication with minor revisions at Philosophy & Technology

♻ ☆ Why Can GPT Learn In-Context? Language Models Secretly Perform Gradient Descent as Meta-Optimizers

Damai Dai, Yutao Sun, Li Dong, Yaru Hao, Zhifang Sui, Furu Wei

Large pretrained language models have shown surprising In-Context Learning (ICL) ability. With a few demonstration input-label pairs, they can predict the label for an unseen input without additional parameter updates. Despite the great success in performance, the working mechanism of ICL still remains an open problem. In order to better understand how ICL works, this paper explains language models as meta-optimizers and understands ICL as a kind of implicit finetuning. Theoretically, we figure out that the Transformer attention has a dual form of gradient descent based optimization. On top of it, we understand ICL as follows: GPT first produces meta-gradients according to the demonstration examples, and then these meta-gradients are applied to the original GPT to build an ICL model. Experimentally, we comprehensively compare the behavior of ICL and explicit finetuning based on real tasks to provide empirical evidence that supports our understanding. The results prove that ICL behaves similarly to explicit finetuning at the prediction level, the representation level, and the attention behavior level. Further, inspired by our understanding of meta-optimization, we design a momentum-based attention by analogy with the momentum-based gradient descent algorithm. Its consistently better performance over vanilla attention supports our understanding again from another aspect, and more importantly, it shows the potential to utilize our understanding for future model designing.

♻ ☆ MMDialog: A Large-scale Multi-turn Dialogue Dataset Towards Multi-modal Open-domain Conversation

Jiazhan Feng, Qingfeng Sun, Can Xu, Pu Zhao, Yaming Yang, Chongyang Tao, Dongyan Zhao, Qingwei Lin

Responding with multi-modal content has been recognized as an essential capability for an intelligent conversational agent. In this paper, we introduce the MMDialog dataset to better facilitate multi-modal conversation. MMDialog is composed of a curated set of 1.08 million real-world dialogues with 1.53 million unique images across 4,184 topics. MMDialog has two main and unique advantages. First, it is the largest multi-modal conversation dataset by the number of dialogues by 88x. Second, it contains massive topics to generalize the open-domain. To build engaging dialogue system with this dataset, we propose and normalize two response producing tasks based on retrieval and generative scenarios. In addition, we build two baselines for above tasks with state-of-the-art techniques and report their experimental performance. We also propose a novel evaluation metric MM-Relevance to measure the multi-modal responses. Our dataset and scripts are available in https://github.com/victorsungo/MMDialog.

♻ ☆ SMSMix: Sense-Maintained Sentence Mixup for Word Sense Disambiguation EMNLP2022

Hee Suk Yoon, Eunseop Yoon, John Harvill, Sunjae Yoon, Mark Hasegawa-Johnson, Chang D. Yoo

Word Sense Disambiguation (WSD) is an NLP task aimed at determining the correct sense of a word in a sentence from discrete sense choices. Although current systems have attained unprecedented performances for such tasks, the nonuniform distribution of word senses during training generally results in systems performing poorly on rare senses. To this end, we consider data augmentation to increase the frequency of these least frequent senses (LFS) to reduce the distributional bias of senses during training. We propose Sense-Maintained Sentence Mixup (SMSMix), a novel word-level mixup method that maintains the sense of a target word. SMSMix smoothly blends two sentences using mask prediction while preserving the relevant span determined by saliency scores to maintain a specific word's sense. To the best of our knowledge, this is the first attempt to apply mixup in NLP while preserving the meaning of a specific word. With extensive experiments, we validate that our augmentation method can effectively give more information about rare senses during training with maintained target sense label.

comment: EMNLP2022

♻ ☆ Two-view Graph Neural Networks for Knowledge Graph Completion

Vinh Tong, Dai Quoc Nguyen, Dinh Phung, Dat Quoc Nguyen

We present an effective graph neural network (GNN)-based knowledge graph embedding model, which we name WGE, to capture entity- and relation-focused graph structures. Given a knowledge graph, WGE builds a single undirected entity-focused graph that views entities as nodes. WGE also constructs another single undirected graph from relation-focused constraints, which views entities and relations as nodes. WGE then proposes a GNN-based architecture to better learn vector representations of entities and relations from these two single entity- and relation-focused graphs. WGE feeds the learned entity and relation representations into a weighted score function to return the triple scores for knowledge graph completion. Experimental results show that WGE outperforms competitive baselines, obtaining state-of-the-art performances on seven benchmark datasets for knowledge graph completion.

comment: 17 pages; 4 tables; 4 figures

♻ ☆ Optimizing Test-Time Query Representations for Dense Retrieval

Mujeen Sung, Jungsoo Park, Jaewoo Kang, Danqi Chen, Jinhyuk Lee

Recent developments of dense retrieval rely on quality representations of queries and contexts coming from pre-trained query and context encoders. In this paper, we introduce TouR (test-time optimization of query representations), which further optimizes instance-level query representations guided by signals from test-time retrieval results. We leverage a cross-encoder re-ranker to provide fine-grained pseudo labels over retrieval results and iteratively optimize query representations with the gradient descent method. Our theoretical analysis reveals that TouR can be viewed as a generalization of the classical Rocchio's algorithm for pseudo relevance feedback, and we present two variants leveraging psuedo labels as either hard binary or soft continuous labels. We first apply TouR on phrase retrieval with our proposed phrase re-ranker. On passage retrieval, we demonstrate its effectiveness with an off-the-shelf re-ranker. TouR improves the end-to-end open-domain QA accuracy significantly, as well as passage retrieval performance. Compared to re-ranker, TouR requires a smaller number of candidates, and achieves consistently better performance and runs up to 4x faster with our efficient implementation.

comment: 15 pages, 5 figures

♻ ☆ Multi-View Active Learning for Short Text Classification in User-Generated Data EMNLP

Payam Karisani, Negin Karisani, Li Xiong

Mining user-generated data often suffers from the lack of enough labeled data, short document lengths, and the informal user language. In this paper, we propose a novel active learning model to overcome these obstacles in the tasks tailored for query phrases--e.g., detecting positive reports of natural disasters. Our model has three novelties: 1) It is the first approach to employ multi-view active learning in this domain. 2) It uses the Parzen-Rosenblatt window method to integrate the representativeness measure into multi-view active learning. 3) It employs a query-by-committee strategy, based on the agreement between predictors, to address the usually noisy language of the documents in this domain. We evaluate our model in four publicly available Twitter datasets with distinctly different applications. We also compare our model with a wide range of baselines including those with multiple classifiers. The experiments testify that our model is highly consistent and outperforms existing models.

comment: EMNLP Findings 2022

♻ ☆ Named Tensor Notation

David Chiang, Alexander M. Rush, Boaz Barak

We propose a notation for tensors with named axes, which relieves the author, reader, and future implementers of machine learning models from the burden of keeping track of the order of axes and the purpose of each. The notation makes it easy to lift operations on low-order tensors to higher order ones, for example, from images to minibatches of images, or from an attention mechanism to multiple attention heads. After a brief overview and formal definition of the notation, we illustrate it through several examples from modern machine learning, from building blocks like attention and convolution to full models like Transformers and LeNet. We then discuss differential calculus in our notation and compare with some alternative notations. Our proposals build on ideas from many previous papers and software libraries. We hope that our notation will encourage more authors to use named tensors, resulting in clearer papers and more precise implementations.

♻ ☆ Reasoning over Different Types of Knowledge Graphs: Static, Temporal and Multi-Modal

Ke Liang, Lingyuan Meng, Meng Liu, Yue Liu, Wenxuan Tu, Siwei Wang, Sihang Zhou, Xinwang Liu, Fuchun Sun

Knowledge graph reasoning (KGR), aiming to deduce new facts from existing facts based on mined logic rules underlying knowledge graphs (KGs), has become a fast-growing research direction. It has been proven to significantly benefit the usage of KGs in many AI applications, such as question answering and recommendation systems, etc. According to the graph types, the existing KGR models can be roughly divided into three categories, i.e., static models, temporal models, and multi-modal models. The early works in this domain mainly focus on static KGR and tend to directly apply general knowledge graph embedding models to the reasoning task. However, these models are not suitable for more complex but practical tasks, such as inductive static KGR, temporal KGR, and multi-modal KGR. To this end, multiple works have been developed recently, but no survey papers and open-source repositories comprehensively summarize and discuss models in this important direction. To fill the gap, we conduct a survey for knowledge graph reasoning tracing from static to temporal and then to multi-modal KGs. Concretely, the preliminaries, summaries of KGR models, and typical datasets are introduced and discussed consequently. Moreover, we discuss the challenges and potential opportunities. The corresponding open-source repository is shared on GitHub: https://github.com/LIANGKE23/Awesome-Knowledge-Graph-Reasoning.

comment: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

☆ Generalized Decoding for Pixel, Image, and Language

Xueyan Zou, Zi-Yi Dou, Jianwei Yang, Zhe Gan, Linjie Li, Chunyuan Li, Xiyang Dai, Harkirat Behl, Jianfeng Wang, Lu Yuan, Nanyun Peng, Lijuan Wang, Yong Jae Lee, Jianfeng Gao

We present X-Decoder, a generalized decoding model that can predict pixel-level segmentation and language tokens seamlessly. X-Decodert takes as input two types of queries: (i) generic non-semantic queries and (ii) semantic queries induced from text inputs, to decode different pixel-level and token-level outputs in the same semantic space. With such a novel design, X-Decoder is the first work that provides a unified way to support all types of image segmentation and a variety of vision-language (VL) tasks. Further, our design enables seamless interactions across tasks at different granularities and brings mutual benefits by learning a common and rich pixel-level visual-semantic understanding space, without any pseudo-labeling. After pretraining on a mixed set of a limited amount of segmentation data and millions of image-text pairs, X-Decoder exhibits strong transferability to a wide range of downstream tasks in both zero-shot and finetuning settings. Notably, it achieves (1) state-of-the-art results on open-vocabulary segmentation and referring segmentation on eight datasets; (2) better or competitive finetuned performance to other generalist and specialist models on segmentation and VL tasks; and (3) flexibility for efficient finetuning and novel task composition (e.g., referring captioning and image editing). Code, demo, video, and visualization are available at https://x-decoder-vl.github.io.

comment: https://x-decoder-vl.github.io

☆ 3D Highlighter: Localizing Regions on 3D Shapes via Text Descriptions

Dale Decatur, Itai Lang, Rana Hanocka

We present 3D Highlighter, a technique for localizing semantic regions on a mesh using text as input. A key feature of our system is the ability to interpret "out-of-domain" localizations. Our system demonstrates the ability to reason about where to place non-obviously related concepts on an input 3D shape, such as adding clothing to a bare 3D animal model. Our method contextualizes the text description using a neural field and colors the corresponding region of the shape using a probability-weighted blend. Our neural optimization is guided by a pre-trained CLIP encoder, which bypasses the need for any 3D datasets or 3D annotations. Thus, 3D Highlighter is highly flexible, general, and capable of producing localizations on a myriad of input shapes. Our code is publicly available at https://github.com/threedle/3DHighlighter.

comment: Project page: https://threedle.github.io/3DHighlighter/

☆ Contrastive Language-Vision AI Models Pretrained on Web-Scraped Multimodal Data Exhibit Sexual Objectification Bias

Robert Wolfe, Yiwei Yang, Bill Howe, Aylin Caliskan

Nine language-vision AI models trained on web scrapes with the Contrastive Language-Image Pretraining (CLIP) objective are evaluated for evidence of a bias studied by psychologists: the sexual objectification of girls and women, which occurs when a person's human characteristics are disregarded and the person is treated as a body or a collection of body parts. A first experiment uses standardized images of women from the Sexual OBjectification and EMotion Database, and finds that, commensurate with prior research in psychology, human characteristics are disassociated from images of objectified women: the model's recognition of emotional state is mediated by whether the subject is fully or partially clothed. Embedding association tests (EATs) return significant effect sizes for both anger (d >.8) and sadness (d >.5). A second experiment measures the effect in a representative application: an automatic image captioner (Antarctic Captions) includes words denoting emotion less than 50% as often for images of partially clothed women than for images of fully clothed women. A third experiment finds that images of female professionals (scientists, doctors, executives) are likely to be associated with sexual descriptions relative to images of male professionals. A fourth experiment shows that a prompt of "a [age] year old girl" generates sexualized images (as determined by an NSFW classifier) up to 73% of the time for VQGAN-CLIP (age 17), and up to 40% of the time for Stable Diffusion (ages 14 and 18); the corresponding rate for boys never surpasses 9%. The evidence indicates that language-vision AI models trained on automatically collected web scrapes learn biases of sexual objectification, which propagate to downstream applications.

comment: 15 pages, 2 figures

☆ Not Just Pretty Pictures: Text-to-Image Generators Enable Interpretable Interventions for Robust Representations

Jianhao Yuan, Francesco Pinto, Adam Davies, Aarushi Gupta, Philip Torr

Neural image classifiers are known to undergo severe performance degradation when exposed to input that exhibits covariate-shift with respect to the training distribution. Successful hand-crafted augmentation pipelines aim at either approximating the expected test domain conditions or to perturb the features that are specific to the training environment. The development of effective pipelines is typically cumbersome, and produce transformations whose impact on the classifier performance are hard to understand and control. In this paper, we show that recent Text-to-Image (T2I) generators' ability to simulate image interventions via natural-language prompts can be leveraged to train more robust models, offering a more interpretable and controllable alternative to traditional augmentation methods. We find that a variety of prompting mechanisms are effective for producing synthetic training data sufficient to achieve state-of-the-art performance in widely-adopted domain-generalization benchmarks and reduce classifiers' dependency on spurious features. Our work suggests that further progress in T2I generation and a tighter integration with other research fields may represent a significant step towards the development of more robust machine learning systems.

☆ Land Cover and Land Use Detection using Semi-Supervised Learning

Fahmida Tasnim Lisa, Md. Zarif Hossain, Sharmin Naj Mou, Shahriar Ivan, Md. Hasanul Kabir

Semi-supervised learning (SSL) has made significant strides in the field of remote sensing. Finding a large number of labeled datasets for SSL methods is uncommon, and manually labeling datasets is expensive and time-consuming. Furthermore, accurately identifying remote sensing satellite images is more complicated than it is for conventional images. Class-imbalanced datasets are another prevalent phenomenon, and models trained on these become biased towards the majority classes. This becomes a critical issue with an SSL model's subpar performance. We aim to address the issue of labeling unlabeled data and also solve the model bias problem due to imbalanced datasets while achieving better accuracy. To accomplish this, we create "artificial" labels and train a model to have reasonable accuracy. We iteratively redistribute the classes through resampling using a distribution alignment technique. We use a variety of class imbalanced satellite image datasets: EuroSAT, UCM, and WHU-RS19. On UCM balanced dataset, our method outperforms previous methods MSMatch and FixMatch by 1.21% and 0.6%, respectively. For imbalanced EuroSAT, our method outperforms MSMatch and FixMatch by 1.08% and 1%, respectively. Our approach significantly lessens the requirement for labeled data, consistently outperforms alternative approaches, and resolves the issue of model bias caused by class imbalance in datasets.

☆ Continual Learning Approaches for Anomaly Detection

Davide Dalle Pezze, Eugenia Anello, Chiara Masiero, Gian Antonio Susto

Anomaly Detection is a relevant problem that arises in numerous real-world applications, especially when dealing with images. However, there has been little research for this task in the Continual Learning setting. In this work, we introduce a novel approach called SCALE (SCALing is Enough) to perform Compressed Replay in a framework for Anomaly Detection in Continual Learning setting. The proposed technique scales and compresses the original images using a Super Resolution model which, to the best of our knowledge, is studied for the first time in the Continual Learning setting. SCALE can achieve a high level of compression while maintaining a high level of image reconstruction quality. In conjunction with other Anomaly Detection approaches, it can achieve optimal results. To validate the proposed approach, we use a real-world dataset of images with pixel-based anomalies, with the scope to provide a reliable benchmark for Anomaly Detection in the context of Continual Learning, serving as a foundation for further advancements in the field.

☆ Similarity Contrastive Estimation for Image and Video Soft Contrastive Self-Supervised Learning WACV 2023

Julien Denize, Jaonary Rabarisoa, Astrid Orcesi, Romain Hérault

Contrastive representation learning has proven to be an effective self-supervised learning method for images and videos. Most successful approaches are based on Noise Contrastive Estimation (NCE) and use different views of an instance as positives that should be contrasted with other instances, called negatives, that are considered as noise. However, several instances in a dataset are drawn from the same distribution and share underlying semantic information. A good data representation should contain relations between the instances, or semantic similarity and dissimilarity, that contrastive learning harms by considering all negatives as noise. To circumvent this issue, we propose a novel formulation of contrastive learning using semantic similarity between instances called Similarity Contrastive Estimation (SCE). Our training objective is a soft contrastive one that brings the positives closer and estimates a continuous distribution to push or pull negative instances based on their learned similarities. We validate empirically our approach on both image and video representation learning. We show that SCE performs competitively with the state of the art on the ImageNet linear evaluation protocol for fewer pretraining epochs and that it generalizes to several downstream image tasks. We also show that SCE reaches state-of-the-art results for pretraining video representation and that the learned representation can generalize to video downstream tasks.

comment: Extended version of our WACV 2023 paper to video self-supervised learning

☆ A recurrent CNN for online object detection on raw radar frames

Colin Decourt, Rufin VanRullen, Didier Salle, Thomas Oberlin

Automotive radar sensors provide valuable information for advanced driving assistance systems (ADAS). Radars can reliably estimate the distance to an object and the relative velocity, regardless of weather and light conditions. However, radar sensors suffer from low resolution and huge intra-class variations in the shape of objects. Exploiting the time information (e.g., multiple frames) has been shown to help to capture better the dynamics of objects and, therefore, the variation in the shape of objects. Most temporal radar object detectors use 3D convolutions to learn spatial and temporal information. However, these methods are often non-causal and unsuitable for real-time applications. This work presents RECORD, a new recurrent CNN architecture for online radar object detection. We propose an end-to-end trainable architecture mixing convolutions and ConvLSTMs to learn spatio-temporal dependencies between successive frames. Our model is causal and requires only the past information encoded in the memory of the ConvLSTMs to detect objects. Our experiments show such a method's relevance for detecting objects in different radar representations (range-Doppler, range-angle) and outperform state-of-the-art models on the ROD2021 and CARRADA datasets while being less computationally expensive. The code will be available soon.

comment: 13 pages, 3 figures

☆ What Makes for Good Tokenizers in Vision Transformer?

Shengju Qian, Yi Zhu, Wenbo Li, Mu Li, Jiaya Jia

The architecture of transformers, which recently witness booming applications in vision tasks, has pivoted against the widespread convolutional paradigm. Relying on the tokenization process that splits inputs into multiple tokens, transformers are capable of extracting their pairwise relationships using self-attention. While being the stemming building block of transformers, what makes for a good tokenizer has not been well understood in computer vision. In this work, we investigate this uncharted problem from an information trade-off perspective. In addition to unifying and understanding existing structural modifications, our derivation leads to better design strategies for vision tokenizers. The proposed Modulation across Tokens (MoTo) incorporates inter-token modeling capability through normalization. Furthermore, a regularization objective TokenProp is embraced in the standard training regime. Through extensive experiments on various transformer architectures, we observe both improved performance and intriguing properties of these two plug-and-play designs with negligible computational overhead. These observations further indicate the importance of the commonly-omitted designs of tokenizers in vision transformer.

comment: To appear in IEEE Transactions on Pattern Analysis and Machine Intelligence

☆ Exploring Content Relationships for Distilling Efficient GANs

Lizhou You, Mingbao Lin, Tie Hu, Fei Chao, Rongrong Ji

This paper proposes a content relationship distillation (CRD) to tackle the over-parameterized generative adversarial networks (GANs) for the serviceability in cutting-edge devices. In contrast to traditional instance-level distillation, we design a novel GAN compression oriented knowledge by slicing the contents of teacher outputs into multiple fine-grained granularities, such as row/column strips (global information) and image patches (local information), modeling the relationships among them, such as pairwise distance and triplet-wise angle, and encouraging the student to capture these relationships within its output contents. Built upon our proposed content-level distillation, we also deploy an online teacher discriminator, which keeps updating when co-trained with the teacher generator and keeps freezing when co-trained with the student generator for better adversarial training. We perform extensive experiments on three benchmark datasets, the results of which show that our CRD reaches the most complexity reduction on GANs while obtaining the best performance in comparison with existing methods. For example, we reduce MACs of CycleGAN by around 40x and parameters by over 80x, meanwhile, 46.61 FIDs are obtained compared with these of 51.92 for the current state-of-the-art. Code of this project is available at https://github.com/TheKernelZ/CRD.

☆ Hi-LASSIE: High-Fidelity Articulated Shape and Skeleton Discovery from Sparse Image Ensemble

Chun-Han Yao, Wei-Chih Hung, Yuanzhen Li, Michael Rubinstein, Ming-Hsuan Yang, Varun Jampani

Automatically estimating 3D skeleton, shape, camera viewpoints, and part articulation from sparse in-the-wild image ensembles is a severely under-constrained and challenging problem. Most prior methods rely on large-scale image datasets, dense temporal correspondence, or human annotations like camera pose, 2D keypoints, and shape templates. We propose Hi-LASSIE, which performs 3D articulated reconstruction from only 20-30 online images in the wild without any user-defined shape or skeleton templates. We follow the recent work of LASSIE that tackles a similar problem setting and make two significant advances. First, instead of relying on a manually annotated 3D skeleton, we automatically estimate a class-specific skeleton from the selected reference image. Second, we improve the shape reconstructions with novel instance-specific optimization strategies that allow reconstructions to faithful fit on each instance while preserving the class-specific priors learned across all images. Experiments on in-the-wild image ensembles show that Hi-LASSIE obtains higher quality state-of-the-art 3D reconstructions despite requiring minimum user input.

☆ Deep set conditioned latent representations for action recognition

Akash Singh, Tom De Schepper, Kevin Mets, Peter Hellinckx, Jose Oramas, Steven Latre

In recent years multi-label, multi-class video action recognition has gained significant popularity. While reasoning over temporally connected atomic actions is mundane for intelligent species, standard artificial neural networks (ANN) still struggle to classify them. In the real world, atomic actions often temporally connect to form more complex composite actions. The challenge lies in recognising composite action of varying durations while other distinct composite or atomic actions occur in the background. Drawing upon the success of relational networks, we propose methods that learn to reason over the semantic concept of objects and actions. We empirically show how ANNs benefit from pretraining, relational inductive biases and unordered set-based latent representations. In this paper we propose deep set conditioned I3D (SCI3D), a two stream relational network that employs latent representation of state and visual representation for reasoning over events and actions. They learn to reason about temporally connected actions in order to identify all of them in the video. The proposed method achieves an improvement of around 1.49% mAP in atomic action recognition and 17.57% mAP in composite action recognition, over a I3D-NL baseline, on the CATER dataset.

comment: Conference VISAPP 2022, 11 pages,5 figures, 2 Tables, 6 plots

☆ Object detection-based inspection of power line insulators: Incipient fault detection in the low data-regime

Laya Das, Mohammad Hossein Saadat, Blazhe Gjorgiev, Etienne Auger, Giovanni Sansavini

Deep learning-based object detection is a powerful approach for detecting faulty insulators in power lines. This involves training an object detection model from scratch, or fine tuning a model that is pre-trained on benchmark computer vision datasets. This approach works well with a large number of insulator images, but can result in unreliable models in the low data regime. The current literature mainly focuses on detecting the presence or absence of insulator caps, which is a relatively easy detection task, and does not consider detection of finer faults such as flashed and broken disks. In this article, we formulate three object detection tasks for insulator and asset inspection from aerial images, focusing on incipient faults in disks. We curate a large reference dataset of insulator images that can be used to learn robust features for detecting healthy and faulty insulators. We study the advantage of using this dataset in the low target data regime by pre-training on the reference dataset followed by fine-tuning on the target dataset. The results suggest that object detection models can be used to detect faults in insulators at a much incipient stage, and that transfer learning adds value depending on the type of object detection model. We identify key factors that dictate performance in the low data-regime and outline potential approaches to improve the state-of-the-art.

☆ Revisiting Residual Networks for Adversarial Robustness: An Architectural Perspective

Shihua Huang, Zhichao Lu, Kalyanmoy Deb, Vishnu Naresh Boddeti

Efforts to improve the adversarial robustness of convolutional neural networks have primarily focused on developing more effective adversarial training methods. In contrast, little attention was devoted to analyzing the role of architectural elements (such as topology, depth, and width) on adversarial robustness. This paper seeks to bridge this gap and present a holistic study on the impact of architectural design on adversarial robustness. We focus on residual networks and consider architecture design at the block level, i.e., topology, kernel size, activation, and normalization, as well as at the network scaling level, i.e., depth and width of each block in the network. In both cases, we first derive insights through systematic ablative experiments. Then we design a robust residual block, dubbed RobustResBlock, and a compound scaling rule, dubbed RobustScaling, to distribute depth and width at the desired FLOP count. Finally, we combine RobustResBlock and RobustScaling and present a portfolio of adversarially robust residual networks, RobustResNets, spanning a broad spectrum of model capacities. Experimental validation across multiple datasets and adversarial attacks demonstrate that RobustResNets consistently outperform both the standard WRNs and other existing robust architectures, achieving state-of-the-art AutoAttack robust accuracy of 61.1% without additional data and 63.7% with 500K external data while being $2\times$ more compact in terms of parameters. Code is available at \url{ https://github.com/zhichao-lu/robust-residual-network}

☆ Attention-Aware Anime Line Drawing Colorization

Yu Cao, Hao Tian, P. Y. Mok

Automatic colorization of anime line drawing has attracted much attention in recent years since it can substantially benefit the animation industry. User-hint based methods are the mainstream approach for line drawing colorization, while reference-based methods offer a more intuitive approach. Nevertheless, although reference-based methods can improve feature aggregation of the reference image and the line drawing, the colorization results are not compelling in terms of color consistency or semantic correspondence. In this paper, we introduce an attention-based model for anime line drawing colorization, in which a channel-wise and spatial-wise Convolutional Attention module is used to improve the ability of the encoder for feature extraction and key area perception, and a Stop-Gradient Attention module with cross-attention and self-attention is used to tackle the cross-domain long-range dependency problem. Extensive experiments show that our method outperforms other SOTA methods, with more accurate line structure and semantic color information.

☆ TruFor: Leveraging all-round clues for trustworthy image forgery detection and localization

Fabrizio Guillaro, Davide Cozzolino, Avneesh Sud, Nicholas Dufour, Luisa Verdoliva

In this paper we present TruFor, a forensic framework that can be applied to a large variety of image manipulation methods, from classic cheapfakes to more recent manipulations based on deep learning. We rely on the extraction of both high-level and low-level traces through a transformer-based fusion architecture that combines the RGB image and a learned noise-sensitive fingerprint. The latter learns to embed the artifacts related to the camera internal and external processing by training only on real data in a self-supervised manner. Forgeries are detected as deviations from the expected regular pattern that characterizes each pristine image. Looking for anomalies makes the approach able to robustly detect a variety of local manipulations, ensuring generalization. In addition to a pixel-level localization map and a whole-image integrity score, our approach outputs a reliability map that highlights areas where localization predictions may be error-prone. This is particularly important in forensic applications in order to reduce false alarms and allow for a large scale analysis. Extensive experiments on several datasets show that our method is able to reliably detect and localize both cheapfakes and deepfakes manipulations outperforming state-of-the-art works. Code will be publicly available at https://grip-unina.github.io/TruFor/

☆ Incremental Learning for Neural Radiance Field with Uncertainty-Filtered Knowledge Distillation

Mengqi Guo, Chen Li, Gim Hee Lee

Recent neural radiance field (NeRF) representation has achieved great success in the tasks of novel view synthesis and 3D reconstruction. However, they suffer from the catastrophic forgetting problem when continuously learning from streaming data without revisiting the previous training data. This limitation prohibits the application of existing NeRF models to scenarios where images come in sequentially. In view of this, we explore the task of incremental learning for neural radiance field representation in this work. We first propose a student-teacher pipeline to mitigate the catastrophic forgetting problem. Specifically, we iterate the process of using the student as the teacher at the end of each incremental step and let the teacher guide the training of the student in the next step. In this way, the student network is able to learn new information from the streaming data and retain old knowledge from the teacher network simultaneously. Given that not all information from the teacher network is helpful since it is only trained with the old data, we further introduce a random inquirer and an uncertainty-based filter to filter useful information. We conduct experiments on the NeRF-synthetic360 and NeRF-real360 datasets, where our approach significantly outperforms the baselines by 7.3% and 25.2% in terms of PSNR. Furthermore, we also show that our approach can be applied to the large-scale camera facing-outwards dataset ScanNet, where we surpass the baseline by 60.0% in PSNR.

☆ Joint Embedding of 2D and 3D Networks for Medical Image Anomaly Detection

Inha Kang, Jinah Park

Obtaining ground truth data in medical imaging has difficulties due to the fact that it requires a lot of annotating time from the experts in the field. Also, when trained with supervised learning, it detects only the cases included in the labels. In real practice, we want to also open to other possibilities than the named cases while examining the medical images. As a solution, the need for anomaly detection that can detect and localize abnormalities by learning the normal characteristics using only normal images is emerging. With medical image data, we can design either 2D or 3D networks of self-supervised learning for anomaly detection task. Although 3D networks, which learns 3D structures of the human body, show good performance in 3D medical image anomaly detection, they cannot be stacked in deeper layers due to memory problems. While 2D networks have advantage in feature detection, they lack 3D context information. In this paper, we develop a method for combining the strength of the 3D network and the strength of the 2D network through joint embedding. We also propose the pretask of self-supervised learning to make it possible for the networks to learn efficiently. Through the experiments, we show that the proposed method achieves better performance in both classification and segmentation tasks compared to the SoTA method.

comment: 15 pages, 11 figures

☆ DCC: A Cascade based Approach to Detect Communities in Social Networks

Soumita Das, Anupam Biswas, Akrati Saxena

Community detection in Social Networks is associated with finding and grouping the most similar nodes inherent in the network. These similar nodes are identified by computing tie strength. Stronger ties indicates higher proximity shared by connected node pairs. This work is motivated by Granovetter's argument that suggests that strong ties lies within densely connected nodes and the theory that community cores in real-world networks are densely connected. In this paper, we have introduced a novel method called \emph{Disjoint Community detection using Cascades (DCC)} which demonstrates the effectiveness of a new local density based tie strength measure on detecting communities. Here, tie strength is utilized to decide the paths followed for propagating information. The idea is to crawl through the tuple information of cascades towards the community core guided by increasing tie strength. Considering the cascade generation step, a novel preferential membership method has been developed to assign community labels to unassigned nodes. The efficacy of $DCC$ has been analyzed based on quality and accuracy on several real-world datasets and baseline community detection algorithms.

comment: To be published in CHSN-2022

☆ A Survey of Mix-based Data Augmentation: Taxonomy, Methods, Applications, and Explainability

Chengtai Cao, Fan Zhou, Yurou Dai, Jianping Wang

Data augmentation (DA) is indispensable in modern machine learning and deep neural networks. The basic idea of DA is to construct new training data to improve the model's generalization by adding slightly disturbed versions of existing data or synthesizing new data. In this work, we review a small but essential subset of DA -- Mix-based Data Augmentation (MixDA) that generates novel samples by mixing multiple examples. Unlike conventional DA approaches based on a single-sample operation or requiring domain knowledge, MixDA is more general in creating a broad spectrum of new data and has received increasing attention in the community. We begin with proposing a new taxonomy classifying MixDA into, Mixup-based, Cutmix-based, and hybrid approaches according to a hierarchical view of the data mix. Various MixDA techniques are then comprehensively reviewed in a more fine-grained way. Owing to its generalization, MixDA has penetrated a variety of applications which are also completely reviewed in this work. We also examine why MixDA works from different aspects of improving model performance, generalization, and calibration while explaining the model behavior based on the properties of MixDA. Finally, we recapitulate the critical findings and fundamental challenges of current MixDA studies, and outline the potential directions for future works. Different from previous related works that summarize the DA approaches in a specific domain (e.g., images or natural language processing) or only review a part of MixDA studies, we are the first to provide a systematical survey of MixDA in terms of its taxonomy, methodology, applications, and explainability. This work can serve as a roadmap to MixDA techniques and application reviews while providing promising directions for researchers interested in this exciting area.

☆ In-Sensor & Neuromorphic Computing are all you need for Energy Efficient Computer Vision

Gourav Datta, Zeyu Liu, Md Abdullah-Al Kaiser, Souvik Kundu, Joe Mathai, Zihan Yin, Ajey P. Jacob, Akhilesh R. Jaiswal, Peter A. Beerel

Due to the high activation sparsity and use of accumulates (AC) instead of expensive multiply-and-accumulates (MAC), neuromorphic spiking neural networks (SNNs) have emerged as a promising low-power alternative to traditional DNNs for several computer vision (CV) applications. However, most existing SNNs require multiple time steps for acceptable inference accuracy, hindering real-time deployment and increasing spiking activity and, consequently, energy consumption. Recent works proposed direct encoding that directly feeds the analog pixel values in the first layer of the SNN in order to significantly reduce the number of time steps. Although the overhead for the first layer MACs with direct encoding is negligible for deep SNNs and the CV processing is efficient using SNNs, the data transfer between the image sensors and the downstream processing costs significant bandwidth and may dominate the total energy. To mitigate this concern, we propose an in-sensor computing hardware-software co-design framework for SNNs targeting image recognition tasks. Our approach reduces the bandwidth between sensing and processing by 12-96x and the resulting total energy by 2.32x compared to traditional CV processing, with a 3.8% reduction in accuracy on ImageNet.

☆ Automatic Network Adaptation for Ultra-Low Uniform-Precision Quantization

Seongmin Park, Beomseok Kwon, Jieun Lim, Kyuyoung Sim, Taeho Kim, Jungwook Choi

Uniform-precision neural network quantization has gained popularity since it simplifies densely packed arithmetic unit for high computing capability. However, it ignores heterogeneous sensitivity to the impact of quantization errors across the layers, resulting in sub-optimal inference accuracy. This work proposes a novel neural architecture search called neural channel expansion that adjusts the network structure to alleviate accuracy degradation from ultra-low uniform-precision quantization. The proposed method selectively expands channels for the quantization sensitive layers while satisfying hardware constraints (e.g., FLOPs, PARAMs). Based on in-depth analysis and experiments, we demonstrate that the proposed method can adapt several popular networks channels to achieve superior 2-bit quantization accuracy on CIFAR10 and ImageNet. In particular, we achieve the best-to-date Top-1/Top-5 accuracy for 2-bit ResNet50 with smaller FLOPs and the parameter size.

☆ TMS-Net: A Segmentation Network Coupled With A Run-time Quality Control Method For Robust Cardiac Image Segmentation

Fatmatulzehra Uslu, Anil A. Bharath

Recently, deep networks have shown impressive performance for the segmentation of cardiac Magnetic Resonance Imaging (MRI) images. However, their achievement is proving slow to transition to widespread use in medical clinics because of robustness issues leading to low trust of clinicians to their results. Predicting run-time quality of segmentation masks can be useful to warn clinicians against poor results. Despite its importance, there are few studies on this problem. To address this gap, we propose a quality control method based on the agreement across decoders of a multi-view network, TMS-Net, measured by the cosine similarity. The network takes three view inputs resliced from the same 3D image along different axes. Different from previous multi-view networks, TMS-Net has a single encoder and three decoders, leading to better noise robustness, segmentation performance and run-time quality estimation in our experiments on the segmentation of the left atrium on STACOM 2013 and STACOM 2018 challenge datasets. We also present a way to generate poor segmentation masks by using noisy images generated with engineered noise and Rician noise to simulate undertraining, high anisotropy and poor imaging settings problems. Our run-time quality estimation method show a good classification of poor and good quality segmentation masks with an AUC reaching to 0.97 on STACOM 2018. We believe that TMS-Net and our run-time quality estimation method has a high potential to increase the thrust of clinicians to automatic image analysis tools.

☆ MoQuad: Motion-focused Quadruple Construction for Video Contrastive Learning ECCV2022

Yuan Liu, Jiacheng Chen, Hao Wu

Learning effective motion features is an essential pursuit of video representation learning. This paper presents a simple yet effective sample construction strategy to boost the learning of motion features in video contrastive learning. The proposed method, dubbed Motion-focused Quadruple Construction (MoQuad), augments the instance discrimination by meticulously disturbing the appearance and motion of both the positive and negative samples to create a quadruple for each video instance, such that the model is encouraged to exploit motion information. Unlike recent approaches that create extra auxiliary tasks for learning motion features or apply explicit temporal modelling, our method keeps the simple and clean contrastive learning paradigm (i.e.,SimCLR) without multi-task learning or extra modelling. In addition, we design two extra training strategies by analyzing initial MoQuad experiments. By simply applying MoQuad to SimCLR, extensive experiments show that we achieve superior performance on downstream tasks compared to the state of the arts. Notably, on the UCF-101 action recognition task, we achieve 93.7% accuracy after pre-training the model on Kinetics-400 for only 200 epochs, surpassing various previous methods

comment: ECCV2022 WorkShop

☆ From Images to Textual Prompts: Zero-shot VQA with Frozen Large Language Models

Jiaxian Guo, Junnan Li, Dongxu Li, Anthony Meng Huat Tiong, Boyang Li, Dacheng Tao, Steven C. H. Hoi

Large language models (LLMs) have demonstrated excellent zero-shot generalization to new language tasks. However, effective utilization of LLMs for zero-shot visual question-answering (VQA) remains challenging, primarily due to the modality disconnection and task disconnection between LLM and VQA task. End-to-end training on vision and language data may bridge the disconnections, but is inflexible and computationally expensive. To address this issue, we propose \emph{Img2Prompt}, a plug-and-play module that provides the prompts that can bridge the aforementioned modality and task disconnections, so that LLMs can perform zero-shot VQA tasks without end-to-end training. In order to provide such prompts, we further employ LLM-agnostic models to provide prompts that can describe image content and self-constructed question-answer pairs, which can effectively guide LLM to perform zero-shot VQA tasks. Img2Prompt offers the following benefits: 1) It can flexibly work with various LLMs to perform VQA. 2)~Without the needing of end-to-end training, it significantly reduces the cost of deploying LLM for zero-shot VQA tasks. 3) It achieves comparable or better performance than methods relying on end-to-end training. For example, we outperform Flamingo~\cite{Deepmind:Flamingo2022} by 5.6\% on VQAv2. On the challenging A-OKVQA dataset, our method even outperforms few-shot methods by as much as 20\%.

☆ Towards Rapid Prototyping and Comparability in Active Learning for Deep Object Detection

Tobias Riedlinger, Marius Schubert, Karsten Kahl, Hanno Gottschalk, Matthias Rottmann

Active learning as a paradigm in deep learning is especially important in applications involving intricate perception tasks such as object detection where labels are difficult and expensive to acquire. Development of active learning methods in such fields is highly computationally expensive and time consuming which obstructs the progression of research and leads to a lack of comparability between methods. In this work, we propose and investigate a sandbox setup for rapid development and transparent evaluation of active learning in deep object detection. Our experiments with commonly used configurations of datasets and detection architectures found in the literature show that results obtained in our sandbox environment are representative of results on standard configurations. The total compute time to obtain results and assess the learning behavior can thereby be reduced by factors of up to 14 when comparing with Pascal VOC and up to 32 when comparing with BDD100k. This allows for testing and evaluating data acquisition and labeling strategies in under half a day and contributes to the transparency and development speed in the field of active learning for object detection.

comment: 17 pages, 12 figures, 9 tables

☆ High-fidelity Direct Contrast Synthesis from Magnetic Resonance Fingerprinting

Ke Wang, Mariya Doneva, Jakob Meineke, Thomas Amthor, Ekin Karasan, Fei Tan, Jonathan I. Tamir, Stella X. Yu, Michael Lustig

Magnetic Resonance Fingerprinting (MRF) is an efficient quantitative MRI technique that can extract important tissue and system parameters such as T1, T2, B0, and B1 from a single scan. This property also makes it attractive for retrospectively synthesizing contrast-weighted images. In general, contrast-weighted images like T1-weighted, T2-weighted, etc., can be synthesized directly from parameter maps through spin-dynamics simulation (i.e., Bloch or Extended Phase Graph models). However, these approaches often exhibit artifacts due to imperfections in the mapping, the sequence modeling, and the data acquisition. Here we propose a supervised learning-based method that directly synthesizes contrast-weighted images from the MRF data without going through the quantitative mapping and spin-dynamics simulation. To implement our direct contrast synthesis (DCS) method, we deploy a conditional Generative Adversarial Network (GAN) framework and propose a multi-branch U-Net as the generator. The input MRF data are used to directly synthesize T1-weighted, T2-weighted, and fluid-attenuated inversion recovery (FLAIR) images through supervised training on paired MRF and target spin echo-based contrast-weighted scans. In-vivo experiments demonstrate excellent image quality compared to simulation-based contrast synthesis and previous DCS methods, both visually as well as by quantitative metrics. We also demonstrate cases where our trained model is able to mitigate in-flow and spiral off-resonance artifacts that are typically seen in MRF reconstructions and thus more faithfully represent conventional spin echo-based contrast-weighted images.

comment: 19 pages, 8 figures

☆ Secure and Privacy Preserving Proxy Biometrics Identities

Harkeerat Kaur, Rishabh Shukla, Isao Echizen, Pritee Khanna

With large-scale adaption to biometric based applications, security and privacy of biometrics is utmost important especially when operating in unsupervised online mode. This work proposes a novel approach for generating new artificial fingerprints also called proxy fingerprints that are natural looking, non-invertible, revocable and privacy preserving. These proxy biometrics can be generated from original ones only with the help of a user-specific key. Instead of using the original fingerprint, these proxy templates can be used anywhere with same convenience. The manuscripts walks through an interesting way in which proxy fingerprints of different types can be generated and how they can be combined with use-specific keys to provide revocability and cancelability in case of compromise. Using the proposed approach a proxy dataset is generated from samples belonging to Anguli fingerprint database. Matching experiments were performed on the new set which is 5 times larger than the original, and it was found that their performance is at par with 0 FAR and 0 FRR in the stolen key, safe key scenarios. Other parameters on revocability and diversity are also analyzed for protection performance.

☆ Semi-Supervised Learning of Monocular Depth Estimation via Consistency Regularization with K-way Disjoint Masking

Jongbeom Baek, Gyeongnyeon Kim, Seonghoon Park, Honggyu An, Matteo Poggi, Seungryong Kim

Semi-Supervised Learning (SSL) has recently accomplished successful achievements in various fields such as image classification, object detection, and semantic segmentation, which typically require a lot of labour to construct ground-truth. Especially in the depth estimation task, annotating training data is very costly and time-consuming, and thus recent SSL regime seems an attractive solution. In this paper, for the first time, we introduce a novel framework for semi-supervised learning of monocular depth estimation networks, using consistency regularization to mitigate the reliance on large ground-truth depth data. We propose a novel data augmentation approach, called K-way disjoint masking, which allows the network for learning how to reconstruct invisible regions so that the model not only becomes robust to perturbations but also generates globally consistent output depth maps. Experiments on the KITTI and NYU-Depth-v2 datasets demonstrate the effectiveness of each component in our pipeline, robustness to the use of fewer and fewer annotated images, and superior results compared to other state-of-the-art, semi-supervised methods for monocular depth estimation. Our code is available at https://github.com/KU-CVLAB/MaskingDepth.

comment: Project page: https://github.com/KU-CVLAB/MaskingDepth

☆ Beyond Information Exchange: An Approach to Deploy Network Properties for Information Diffusion

Soumita Das, Anupam Biswas, Ravi Kishore Devarapalli

Information diffusion in Online Social Networks is a new and crucial problem in social network analysis field and requires significant research attention. Efficient diffusion of information are of critical importance in diverse situations such as; pandemic prevention, advertising, marketing etc. Although several mathematical models have been developed till date, but previous works lacked systematic analysis and exploration of the influence of neighborhood for information diffusion. In this paper, we have proposed Common Neighborhood Strategy (CNS) algorithm for information diffusion that demonstrates the role of common neighborhood in information propagation throughout the network. The performance of CNS algorithm is evaluated on several real-world datasets in terms of diffusion speed and diffusion outspread and compared with several widely used information diffusion models. Empirical results show CNS algorithm enables better information diffusion both in terms of diffusion speed and diffusion outspread.

comment: To be published in BigDML 2021

☆ Direct Comparative Analysis of Nature-inspired Optimization Algorithms on Community Detection Problem in Social Networks

Soumita Das, Bijita Singha, Alberto Tonda, Anupam Biswas

Nature-inspired optimization Algorithms (NIOAs) are nowadays a popular choice for community detection in social networks. Community detection problem in social network is treated as optimization problem, where the objective is to either maximize the connection within the community or minimize connections between the communities. To apply NIOAs, either of the two, or both objectives are explored. Since NIOAs mostly exploit randomness in their strategies, it is necessary to analyze their performance for specific applications. In this paper, NIOAs are analyzed on the community detection problem. A direct comparison approach is followed to perform pairwise comparison of NIOAs. The performance is measured in terms of five scores designed based on prasatul matrix and also with average isolability. Three widely used real-world social networks and four NIOAs are considered for analyzing the quality of communities generated by NIOAs.

☆ Low-Light Image and Video Enhancement: A Comprehensive Survey and Beyond

Shen Zheng, Yiling Ma, Jinqian Pan, Changjie Lu, Gaurav Gupta

This paper presents a comprehensive survey of low-light image and video enhancement. We begin with the challenging mixed over-/under-exposed images, which are under-performed by existing methods. To this end, we propose two variants of the SICE dataset named SICE_Grad and SICE_Mix. Next, we introduce Night Wenzhou, a large-scale, high-resolution video dataset, to address the issue of the lack of a low-light video dataset that discount the use of low-light image enhancement (LLIE) to videos. The Night Wenzhou dataset is challenging since it consists of fast-moving aerial scenes and streetscapes with varying illuminations and degradation. We conduct extensive key technique analysis and experimental comparisons for representative LLIE approaches using these newly proposed datasets and the current benchmark datasets. Finally, we address unresolved issues and propose future research topics for the LLIE community.

comment: 10 pages, 8 tables, and 13 figures

☆ Class Prototype-based Cleaner for Label Noise Learning

Jingjia Huang, Yuanqi Chen, Jiashi Feng, Xinglong Wu

Semi-supervised learning based methods are current SOTA solutions to the noisy-label learning problem, which rely on learning an unsupervised label cleaner first to divide the training samples into a labeled set for clean data and an unlabeled set for noise data. Typically, the cleaner is obtained via fitting a mixture model to the distribution of per-sample training losses. However, the modeling procedure is \emph{class agnostic} and assumes the loss distributions of clean and noise samples are the same across different classes. Unfortunately, in practice, such an assumption does not always hold due to the varying learning difficulty of different classes, thus leading to sub-optimal label noise partition criteria. In this work, we reveal this long-ignored problem and propose a simple yet effective solution, named \textbf{C}lass \textbf{P}rototype-based label noise \textbf{C}leaner (\textbf{CPC}). Unlike previous works treating all the classes equally, CPC fully considers loss distribution heterogeneity and applies class-aware modulation to partition the clean and noise data. CPC takes advantage of loss distribution modeling and intra-class consistency regularization in feature space simultaneously and thus can better distinguish clean and noise labels. We theoretically justify the effectiveness of our method by explaining it from the Expectation-Maximization (EM) framework. Extensive experiments are conducted on the noisy-label benchmarks CIFAR-10, CIFAR-100, Clothing1M and WebVision. The results show that CPC consistently brings about performance improvement across all benchmarks. Codes and pre-trained models will be released at \url{https://github.com/hjjpku/CPC.git}.

☆ SLGTformer: An Attention-Based Approach to Sign Language Recognition

Neil Song

Sign language is the preferred method of communication of deaf or mute people, but similar to any language, it is difficult to learn and represents a significant barrier for those who are hard of hearing or unable to speak. A person's entire frontal appearance dictates and conveys specific meaning. However, this frontal appearance can be quantified as a temporal sequence of human body pose, leading to Sign Language Recognition through the learning of spatiotemporal dynamics of skeleton keypoints. I propose a novel, attention-based approach to Sign Language Recognition exclusively built upon decoupled graph and temporal self-attention: the Sign Language Graph Time Transformer (SLGTformer). SLGTformer first deconstructs spatiotemporal pose sequences separately into spatial graphs and temporal windows. SLGTformer then leverages novel Learnable Graph Relative Positional Encodings (LGRPE) to guide spatial self-attention with the graph neighborhood context of the human skeleton. By modeling the temporal dimension as intra- and inter-window dynamics, I introduce Temporal Twin Self-Attention (TTSA) as the combination of locally-grouped temporal attention (LTA) and global sub-sampled temporal attention (GSTA). I demonstrate the effectiveness of SLGTformer on the World-Level American Sign Language (WLASL) dataset, achieving state-of-the-art performance with an ensemble-free approach on the keypoint modality.

comment: 12 pages, 3 figures, The code is available at \url{https://github.com/neilsong/slt}

☆ An Audio-Visual Speech Separation Model Inspired by Cortico-Thalamo-Cortical Circuits

Kai Li, Fenghua Xie, Hang Chen, Kexin Yuan, Xiaolin Hu

Audio-visual approaches involving visual inputs have laid the foundation for recent progress in speech separation. However, the optimization of the concurrent usage of auditory and visual inputs is still an active research area. Inspired by the cortico-thalamo-cortical circuit, in which the sensory processing mechanisms of different modalities modulate one another via the non-lemniscal sensory thalamus, we propose a novel cortico-thalamo-cortical neural network (CTCNet) for audio-visual speech separation (AVSS). First, the CTCNet learns hierarchical auditory and visual representations in a bottom-up manner in separate auditory and visual subnetworks, mimicking the functions of the auditory and visual cortical areas. Then, inspired by the large number of connections between cortical regions and the thalamus, the model fuses the auditory and visual information in a thalamic subnetwork through top-down connections. Finally, the model transmits this fused information back to the auditory and visual subnetworks, and the above process is repeated several times. The results of experiments on three speech separation benchmark datasets show that CTCNet remarkably outperforms existing AVSS methods with considerablely fewer parameters. These results suggest that mimicking the anatomical connectome of the mammalian brain has great potential for advancing the development of deep neural networks. Project repo is https://github.com/JusperLee/CTCNet.

comment: 13 pages, 5 figures

☆ NADBenchmarks -- a compilation of Benchmark Datasets for Machine Learning Tasks related to Natural Disasters

Adiba Mahbub Proma, Md Saiful Islam, Stela Ciko, Raiyan Abdul Baten, Ehsan Hoque

Climate change has increased the intensity, frequency, and duration of extreme weather events and natural disasters across the world. While the increased data on natural disasters improves the scope of machine learning (ML) in this field, progress is relatively slow. One bottleneck is the lack of benchmark datasets that would allow ML researchers to quantify their progress against a standard metric. The objective of this short paper is to explore the state of benchmark datasets for ML tasks related to natural disasters, categorizing them according to the disaster management cycle. We compile a list of existing benchmark datasets introduced in the past five years. We propose a web platform - NADBenchmarks - where researchers can search for benchmark datasets for natural disasters, and we develop a preliminary version of such a platform using our compiled list. This paper is intended to aid researchers in finding benchmark datasets to train their ML models on, and provide general directions for topics where they can contribute new benchmark datasets.

☆ UnICLAM:Contrastive Representation Learning with Adversarial Masking for Unified and Interpretable Medical Vision Question Answering

Chenlu Zhan, Peng Peng, Hongsen Wang, Tao Chen, Hongwei Wang

Medical Visual Question Answering (Medical-VQA) aims to answer clinical questions regarding radiology images, assisting doctors with decision-making options. Nevertheless, current Medical-VQA models learn cross-modal representations through residing vision and texture encoders in dual separate spaces, which lead to indirect semantic alignment. In this paper, we propose UnICLAM, a Unified and Interpretable Medical-VQA model through Contrastive Representation Learning with Adversarial Masking. Specifically, to learn an aligned image-text representation, we first establish a unified dual-stream pre-training structure with the gradually soft-parameter sharing strategy. Technically, the proposed strategy learns a constraint for the vision and texture encoders to be close in a same space, which is gradually loosened as the higher number of layers. Moreover, for grasping the semantic representation, we extend the unified Adversarial Masking data augmentation strategy to the contrastive representation learning of vision and text in a unified manner, alleviating the meaningless of the commonly used random mask. Concretely, while the encoder training minimizes the distance between the original feature and the masking feature, the adversarial masking model keeps adversarial learning to conversely maximize the distance. Furthermore, we also intuitively take a further exploration of the unified adversarial masking strategy, which improves the potential ante-hoc interpretability with remarkable performance and efficiency. Experimental results on VQA-RAD and SLAKE public benchmarks demonstrate that UnICLAM outperforms the existing 11 state-of-the-art Medical-VQA models. More importantly, we make an additional discussion about the performance of UnICLAM in diagnosing heart failure, verifying that UnICLAM exhibits superior few-shot adaption performance in practical disease diagnosis.

☆ Investigation of Network Architecture for Multimodal Head-and-Neck Tumor Segmentation

Ye Li, Junyu Chen, Se-in Jang, Kuang Gong, Quanzheng Li

Inspired by the recent success of Transformers for Natural Language Processing and vision Transformer for Computer Vision, many researchers in the medical imaging community have flocked to Transformer-based networks for various main stream medical tasks such as classification, segmentation, and estimation. In this study, we analyze, two recently published Transformer-based network architectures for the task of multimodal head-and-tumor segmentation and compare their performance to the de facto standard 3D segmentation network - the nnU-Net. Our results showed that modeling long-range dependencies may be helpful in cases where large structures are present and/or large field of view is needed. However, for small structures such as head-and-neck tumor, the convolution-based U-Net architecture seemed to perform well, especially when training dataset is small and computational resource is limited.

comment: Accepted for oral presentation by IEEE Medical Imaging Conference 2022

☆ PaletteNeRF: Palette-based Appearance Editing of Neural Radiance Fields

Zhengfei Kuang, Fujun Luan, Sai Bi, Zhixin Shu, Gordon Wetzstein, Kalyan Sunkavalli

Recent advances in neural radiance fields have enabled the high-fidelity 3D reconstruction of complex scenes for novel view synthesis. However, it remains underexplored how the appearance of such representations can be efficiently edited while maintaining photorealism. In this work, we present PaletteNeRF, a novel method for photorealistic appearance editing of neural radiance fields (NeRF) based on 3D color decomposition. Our method decomposes the appearance of each 3D point into a linear combination of palette-based bases (i.e., 3D segmentations defined by a group of NeRF-type functions) that are shared across the scene. While our palette-based bases are view-independent, we also predict a view-dependent function to capture the color residual (e.g., specular shading). During training, we jointly optimize the basis functions and the color palettes, and we also introduce novel regularizers to encourage the spatial coherence of the decomposition. Our method allows users to efficiently edit the appearance of the 3D scene by modifying the color palettes. We also extend our framework with compressed semantic features for semantic-aware appearance editing. We demonstrate that our technique is superior to baseline methods both quantitatively and qualitatively for appearance editing of complex real-world scenes.

☆ Cattle Detection Occlusion Problem

Aparna Mendu, Bhavya Sehgal, Vaishnavi Mendu

The management of cattle over a huge area is still a challenging problem in the farming sector. With evolution in technology, Unmanned aerial vehicles (UAVs) with consumer level digital cameras are becoming a popular alternative to manual animal censuses for livestock estimation since they are less risky and expensive.This paper evaluated and compared the cutting-edge object detection algorithms, YOLOv7,RetinaNet with ResNet50 backbone, RetinaNet with EfficientNet and mask RCNN. It aims to improve the occlusion problem that is to detect hidden cattle from a huge dataset captured by drones using deep learning algorithms for accurate cattle detection. Experimental results showed YOLOv7 was superior with precision of 0.612 when compared to the other two algorithms. The proposed method proved superior to the usual competing algorithms for cow face detection, especially in very difficult cases.

comment: arXiv admin note: text overlap with arXiv:1701.01611 by other authors

☆ DExT: Detector Explanation Toolkit

Deepan Chakravarthi Padmanabhan, Matias Valdenegro-Toro

State-of-the-art object detectors are treated as black boxes due to their highly non-linear internal computations. Even with unprecedented advancements in detector performance, the inability to explain how their outputs are generated limits their use in safety-critical applications. Previous work fails to produce explanations for both bounding box and classification decisions, and generally make individual explanations for various detectors. In this paper, we propose an open-source Detector Explanation Toolkit (DExT) which implements the proposed approach to generate a holistic explanation for all detector decisions using certain gradient-based explanation methods. We suggests various multi-object visualization methods to merge the explanations of multiple objects detected in an image as well as the corresponding detections in a single image. The quantitative evaluation show that the Single Shot MultiBox Detector (SSD) is more faithfully explained compared to other detectors regardless of the explanation methods. Both quantitative and human-centric evaluations identify that SmoothGrad with Guided Backpropagation (GBP) provides more trustworthy explanations among selected methods across all detectors. We expect that DExT will motivate practitioners to evaluate object detectors from the interpretability perspective by explaining both bounding box and classification decisions.

comment: 21 pages, with supplementary

☆ ReVISE: Self-Supervised Speech Resynthesis with Visual Input for Universal and Generalized Speech Enhancement

Wei-Ning Hsu, Tal Remez, Bowen Shi, Jacob Donley, Yossi Adi

Prior works on improving speech quality with visual input typically study each type of auditory distortion separately (e.g., separation, inpainting, video-to-speech) and present tailored algorithms. This paper proposes to unify these subjects and study Generalized Speech Enhancement, where the goal is not to reconstruct the exact reference clean signal, but to focus on improving certain aspects of speech. In particular, this paper concerns intelligibility, quality, and video synchronization. We cast the problem as audio-visual speech resynthesis, which is composed of two steps: pseudo audio-visual speech recognition (P-AVSR) and pseudo text-to-speech synthesis (P-TTS). P-AVSR and P-TTS are connected by discrete units derived from a self-supervised speech model. Moreover, we utilize self-supervised audio-visual speech model to initialize P-AVSR. The proposed model is coined ReVISE. ReVISE is the first high-quality model for in-the-wild video-to-speech synthesis and achieves superior performance on all LRS3 audio-visual enhancement tasks with a single model. To demonstrates its applicability in the real world, ReVISE is also evaluated on EasyCom, an audio-visual benchmark collected under challenging acoustic conditions with only 1.6 hours of training data. Similarly, ReVISE greatly suppresses noise and improves quality. Project page: https://wnhsu.github.io/ReVISE.

☆ Artistic Arbitrary Style Transfer

Weiting Li, Rahul Vyas, Ramya Sree Penta

Arbitrary Style Transfer is a technique used to produce a new image from two images: a content image, and a style image. The newly produced image is unseen and is generated from the algorithm itself. Balancing the structure and style components has been the major challenge that other state-of-the-art algorithms have tried to solve. Despite all the efforts, it's still a major challenge to apply the artistic style that was originally created on top of the structure of the content image while maintaining consistency. In this work, we solved these problems by using a Deep Learning approach using Convolutional Neural Networks. Our implementation will first extract foreground from the background using the pre-trained Detectron 2 model from the content image, and then apply the Arbitrary Style Transfer technique that is used in SANet. Once we have the two styled images, we will stitch the two chunks of images after the process of style transfer for the complete end piece.

comment: 12 pages, 7 figures

☆ Semi-supervised GAN for Bladder Tissue Classification in Multi-Domain Endoscopic Images

Jorge F. Lazo, Benoit Rosa, Michele Catellani, Matteo Fontana, Francesco A. Mistretta, Gennaro Musi, Ottavio de Cobelli, Michel de Mathelin, Elena De Momi

Objective: Accurate visual classification of bladder tissue during Trans-Urethral Resection of Bladder Tumor (TURBT) procedures is essential to improve early cancer diagnosis and treatment. During TURBT interventions, White Light Imaging (WLI) and Narrow Band Imaging (NBI) techniques are used for lesion detection. Each imaging technique provides diverse visual information that allows clinicians to identify and classify cancerous lesions. Computer vision methods that use both imaging techniques could improve endoscopic diagnosis. We address the challenge of tissue classification when annotations are available only in one domain, in our case WLI, and the endoscopic images correspond to an unpaired dataset, i.e. there is no exact equivalent for every image in both NBI and WLI domains. Method: We propose a semi-surprised Generative Adversarial Network (GAN)-based method composed of three main components: a teacher network trained on the labeled WLI data; a cycle-consistency GAN to perform unpaired image-to-image translation, and a multi-input student network. To ensure the quality of the synthetic images generated by the proposed GAN we perform a detailed quantitative, and qualitative analysis with the help of specialists. Conclusion: The overall average classification accuracy, precision, and recall obtained with the proposed method for tissue classification are 0.90, 0.88, and 0.89 respectively, while the same metrics obtained in the unlabeled domain (NBI) are 0.92, 0.64, and 0.94 respectively. The quality of the generated images is reliable enough to deceive specialists. Significance: This study shows the potential of using semi-supervised GAN-based classification to improve bladder tissue classification when annotations are limited in multi-domain data.

☆ MM811 Project Report: Cloud Detection and Removal in Satellite Images

Dale Chen-Song, Erfan Khalaji, Vaishali Rani

For satellite images, the presence of clouds presents a problem as clouds obscure more than half to two-thirds of the ground information. This problem causes many issues for reliability in a noise-free environment to communicate data and other applications that need seamless monitoring. Removing the clouds from the images while keeping the background pixels intact can help address the mentioned issues. Recently, deep learning methods have become popular for researching cloud removal by demonstrating promising results, among which Generative Adversarial Networks (GAN) have shown considerably better performance. In this project, we aim to address cloud removal from satellite images using AttentionGAN and then compare our results by reproducing the results obtained using traditional GANs and auto-encoders. We use RICE dataset. The outcome of this project can be used to develop applications that require cloud-free satellite images. Moreover, our results could be helpful for making further research improvements.

☆ Lightweight Monocular Depth Estimation

Ruilin Ma, Shiyao Chen, Qin Zhang

Monocular depth estimation can play an important role in addressing the issue of deriving scene geometry from 2D images. It has been used in a variety of industries, including robots, self-driving cars, scene comprehension, 3D reconstructions, and others. The goal of our method is to create a lightweight machine-learning model in order to predict the depth value of each pixel given only a single RGB image as input with the Unet structure of the image segmentation network. We use the NYU Depth V2 dataset to test the structure and compare the result with other methods. The proposed method achieves relatively high accuracy and low rootmean-square error.

☆ Knowledge-driven Scene Priors for Semantic Audio-Visual Embodied Navigation

Gyan Tatiya, Jonathan Francis, Luca Bondi, Ingrid Navarro, Eric Nyberg, Jivko Sinapov, Jean Oh

Generalisation to unseen contexts remains a challenge for embodied navigation agents. In the context of semantic audio-visual navigation (SAVi) tasks, the notion of generalisation should include both generalising to unseen indoor visual scenes as well as generalising to unheard sounding objects. However, previous SAVi task definitions do not include evaluation conditions on truly novel sounding objects, resorting instead to evaluating agents on unheard sound clips of known objects; meanwhile, previous SAVi methods do not include explicit mechanisms for incorporating domain knowledge about object and region semantics. These weaknesses limit the development and assessment of models' abilities to generalise their learned experience. In this work, we introduce the use of knowledge-driven scene priors in the semantic audio-visual embodied navigation task: we combine semantic information from our novel knowledge graph that encodes object-region relations, spatial knowledge from dual Graph Encoder Networks, and background knowledge from a series of pre-training tasks -- all within a reinforcement learning framework for audio-visual navigation. We also define a new audio-visual navigation sub-task, where agents are evaluated on novel sounding objects, as opposed to unheard clips of known objects. We show improvements over strong baselines in generalisation to unseen regions and novel sounding objects, within the Habitat-Matterport3D simulation environment, under the SoundSpaces task.

comment: 19 pages, 8 figures, 9 tables

☆ Advanced Baseline for 3D Human Pose Estimation: A Two-Stage Approach

Zichen Gui, Jungang Luo

Human pose estimation has been widely applied in various industries. While recent decades have witnessed the introduction of many advanced two-dimensional (2D) human pose estimation solutions, three-dimensional (3D) human pose estimation is still an active research field in computer vision. Generally speaking, 3D human pose estimation methods can be divided into two categories: single-stage and two-stage. In this paper, we focused on the 2D-to-3D lifting process in the two-stage methods and proposed a more advanced baseline model for 3D human pose estimation, based on the existing solutions. Our improvements include optimization of machine learning models and multiple parameters, as well as introduction of a weighted loss to the training model. Finally, we used the Human3.6M benchmark to test the final performance and it did produce satisfactory results.

☆ DuAT: Dual-Aggregation Transformer Network for Medical Image Segmentation

Feilong Tang, Qiming Huang, Jinfeng Wang, Xianxu Hou, Jionglong Su, Jingxin Liu

Transformer-based models have been widely demonstrated to be successful in computer vision tasks by modelling long-range dependencies and capturing global representations. However, they are often dominated by features of large patterns leading to the loss of local details (e.g., boundaries and small objects), which are critical in medical image segmentation. To alleviate this problem, we propose a Dual-Aggregation Transformer Network called DuAT, which is characterized by two innovative designs, namely, the Global-to-Local Spatial Aggregation (GLSA) and Selective Boundary Aggregation (SBA) modules. The GLSA has the ability to aggregate and represent both global and local spatial features, which are beneficial for locating large and small objects, respectively. The SBA module is used to aggregate the boundary characteristic from low-level features and semantic information from high-level features for better preserving boundary details and locating the re-calibration objects. Extensive experiments in six benchmark datasets demonstrate that our proposed model outperforms state-of-the-art methods in the segmentation of skin lesion images, and polyps in colonoscopy images. In addition, our approach is more robust than existing methods in various challenging situations such as small object segmentation and ambiguous object boundaries.

♻ ☆ On the Cross-dataset Generalization in License Plate Recognition

Rayson Laroca, Everton V. Cardoso, Diego R. Lucio, Valter Estevam, David Menotti

Automatic License Plate Recognition (ALPR) systems have shown remarkable performance on license plates (LPs) from multiple regions due to advances in deep learning and the increasing availability of datasets. The evaluation of deep ALPR systems is usually done within each dataset; therefore, it is questionable if such results are a reliable indicator of generalization ability. In this paper, we propose a traditional-split versus leave-one-dataset-out experimental setup to empirically assess the cross-dataset generalization of 12 Optical Character Recognition (OCR) models applied to LP recognition on nine publicly available datasets with a great variety in several aspects (e.g., acquisition settings, image resolution, and LP layouts). We also introduce a public dataset for end-to-end ALPR that is the first to contain images of vehicles with Mercosur LPs and the one with the highest number of motorcycle images. The experimental results shed light on the limitations of the traditional-split protocol for evaluating approaches in the ALPR context, as there are significant drops in performance for most datasets when training and testing the models in a leave-one-dataset-out fashion.

comment: Accepted for presentation at the International Conference on Computer Vision Theory and Applications (VISAPP) 2022

♻ ☆ FastBVP-Net: a lightweight pulse extraction network for measuring heart rhythm via facial videos

Jialiang Zhuang, Yuheng Chen, Yun Zhang, Xiujuan Zheng

Remote photoplethysmography (rPPG) is an attractive camera-based health monitoring method that can measure the heart rhythm from facial videos. Many well-established deep-learning models have been reported to measure heart rate (HR) and heart rate variability (HRV). However, most of these models usually require a 30-second facial video and enormous computational resources to obtain accurate and robust results, which significantly limits their applications in real-world scenarios. Hence, we propose a lightweight pulse extraction network, FastBVP-Net, to quickly measure heart rhythm via facial videos. The proposed FastBVP-Net uses a multi-frequency mode signal fusion (MMSF) mechanism to characterize the different modes of the raw signals in a decompose module and reconstruct the blood volume pulse (BVP) signal under a complex noise environment in a compose module. Meanwhile, an oversampling training scheme is used to solve the over-fitting problem caused by the limitations of the datasets. Then, the HR and HRV can be estimated based on the extracted BVP signals. Comprehensive experiments are conducted on the benchmark datasets to validate the proposed FastBVP-Net. For intra-dataset and cross-dataset testing, the proposed approach achieves better performance for HR and HRV estimation from 30-second facial videos with fewer computational burdens than the current well-established methods. Moreover, the proposed approach also achieves competitive results from 15-second facial videos. Therefore, the proposed FastBVP-Net has the potential to be applied in many real-world scenarios with shorter videos.

comment: 9 pages, 2figures

♻ ☆ Deconstructing Self-Supervised Monocular Reconstruction: The Design Decisions that Matter

Jaime Spencer, Chris Russell, Simon Hadfield, Richard Bowden

This paper presents an open and comprehensive framework to systematically evaluate state-of-the-art contributions to self-supervised monocular depth estimation. This includes pretraining, backbone, architectural design choices and loss functions. Many papers in this field claim novelty in either architecture design or loss formulation. However, simply updating the backbone of historical systems results in relative improvements of 25%, allowing them to outperform the majority of existing systems. A systematic evaluation of papers in this field was not straightforward. The need to compare like-with-like in previous papers means that longstanding errors in the evaluation protocol are ubiquitous in the field. It is likely that many papers were not only optimized for particular datasets, but also for errors in the data and evaluation criteria. To aid future research in this area, we release a modular codebase (https://github.com/jspenmar/monodepth_benchmark), allowing for easy evaluation of alternate design decisions against corrected data and evaluation criteria. We re-implement, validate and re-evaluate 16 state-of-the-art contributions and introduce a new dataset (SYNS-Patches) containing dense outdoor depth maps in a variety of both natural and urban scenes. This allows for the computation of informative metrics in complex regions such as depth boundaries.

comment: https://github.com/jspenmar/monodepth_benchmark

♻ ☆ MUS-CDB: Mixed Uncertainty Sampling with Class Distribution Balancing for Active Annotation in Aerial Object Detection

Dong Liang, Jing-Wei Zhang, Ying-Peng Tang, Sheng-Jun Huang

Recent aerial object detection models rely on a large amount of labeled training data, which requires unaffordable manual labeling costs in large aerial scenes with dense objects. Active learning is effective in reducing the data labeling cost by selectively querying the informative and representative unlabelled samples. However, existing active learning methods are mainly with class-balanced setting and image-based querying for generic object detection tasks, which are less applicable to aerial object detection scenario due to the long-tailed class distribution and dense small objects in aerial scenes. In this paper, we propose a novel active learning method for cost-effective aerial object detection. Specifically, both object-level and image-level informativeness are considered in the object selection to refrain from redundant and myopic querying. Besides, an easy-to-use class-balancing criterion is incorporated to favor the minority objects to alleviate the long-tailed class distribution problem in model training. To fully utilize the queried information, we further devise a training loss to mine the latent knowledge in the undiscovered image regions. Extensive experiments are conducted on the DOTA-v1.0 and DOTA-v2.0 benchmarks to validate the effectiveness of the proposed method. The results show that it can save more than 75% of the labeling cost to reach the same performance compared to the baselines and state-of-the-art active object detection methods. Code is available at \href{https://github.com/ZJW700/MUS-CDB}{\textit{https://github.com/ZJW700/MUS-CDB}}.

comment: 13 pages, 7 figures

♻ ☆ Diversity-Promoting Ensemble for Medical Image Segmentation

Mariana-Iuliana Georgescu, Radu Tudor Ionescu, Andreea-Iuliana Miron

Medical image segmentation is an actively studied task in medical imaging, where the precision of the annotations is of utter importance towards accurate diagnosis and treatment. In recent years, the task has been approached with various deep learning systems, among the most popular models being U-Net. In this work, we propose a novel strategy to generate ensembles of different architectures for medical image segmentation, by leveraging the diversity (decorrelation) of the models forming the ensemble. More specifically, we utilize the Dice score among model pairs to estimate the correlation between the outputs of the two models forming each pair. To promote diversity, we select models with low Dice scores among each other. We carry out gastro-intestinal tract image segmentation experiments to compare our diversity-promoting ensemble (DiPE) with another strategy to create ensembles based on selecting the top scoring U-Net models. Our empirical results show that DiPE surpasses both individual models as well as the ensemble creation strategy based on selecting the top scoring models.

comment: Accepted at SAC 2023

♻ ☆ Classification and mapping of low-statured 'shrubland' cover types in post-agricultural landscapes of the US Northeast

Michael J Mahoney, Lucas K Johnson, Abigail Z Guinan, Colin M Beier

Novel plant communities reshape landscapes and pose challenges for land cover classification and mapping that can constrain research and stewardship efforts. In the US Northeast, emergence of low-statured woody vegetation, or shrublands, instead of secondary forests in post-agricultural landscapes is well-documented by field studies, but poorly understood from a landscape perspective, which limits the ability to systematically study and manage these lands. To address gaps in classification/mapping of low-statured cover types where they have been historically rare, we developed models to predict shrubland distributions at 30m resolution across New York State (NYS), using a stacked ensemble combining a random forest, gradient boosting machine, and artificial neural network to integrate remote sensing of structural (airborne LIDAR) and optical (satellite imagery) properties of vegetation cover. We first classified a 1m canopy height model (CHM), derived from a patchwork of available LIDAR coverages, to define shrubland presence/absence. Next, these non-contiguous maps were used to train a model ensemble based on temporally-segmented imagery to predict shrubland probability for the entire study landscape (NYS). Approximately 2.5% of the CHM coverage area was classified as shrubland. Models using Landsat predictors trained on the classified CHM were effective at identifying shrubland (test set AUC=0.893, real-world AUC=0.904), in discriminating between shrub/young forest and other cover classes, and produced qualitatively sensible maps, even when extending beyond the original training data. Our results suggest that incorporation of airborne LiDAR, even from a discontinuous patchwork of coverages, can improve land cover classification of historically rare but increasingly prevalent shrubland habitats across broader areas.

comment: 43 pages (35 main text, 8 supplementary materials); 11 figures (10 main text, 1 supplementary materials), 10 tables (4 main text, 6 supplementary materials)

♻ ☆ Image quality prediction using synthetic and natural codebooks: comparative results

Maxim Koroteev, Kirill Aistov, Valeriy Berezovskiy, Pavel Frolov

We investigate a model for image/video quality assessment based on building a set of codevectors representing in a sense some basic properties of images, similar to well-known CORNIA model. We analyze the codebook building method and propose some modifications for it. Also the algorithm is investigated from the point of inference time reduction. Both natural and synthetic images are used for building codebooks and some analysis of synthetic images used for codebooks is provided. It is demonstrated the results on quality assessment may be improves with the use if synthetic images for codebook construction. We also demonstrate regimes of the algorithm in which real time execution on CPU is possible for sufficiently high correlations with mean opinion score (MOS). Various pooling strategies are considered as well as the problem of metric sensitivity to bitrate.

comment: 18 pages, 8 figures

♻ ☆ Automated analysis of diabetic retinopathy using vessel segmentation maps as inductive bias MICCAI 2022

Linus Kreitner, Ivan Ezhov, Daniel Rueckert, Johannes C. Paetzold, Martin J. Menten

Recent studies suggest that early stages of diabetic retinopathy (DR) can be diagnosed by monitoring vascular changes in the deep vascular complex. In this work, we investigate a novel method for automated DR grading based on optical coherence tomography angiography (OCTA) images. Our work combines OCTA scans with their vessel segmentations, which then serve as inputs to task specific networks for lesion segmentation, image quality assessment and DR grading. For this, we generate synthetic OCTA images to train a segmentation network that can be directly applied on real OCTA data. We test our approach on MICCAI 2022's DR analysis challenge (DRAC). In our experiments, the proposed method performs equally well as the baseline model.

comment: Submission for MICCAI 2022 Diabetic Retinopathy Analysis Challenge (DRAC) Proceedings, DOI: 10.5281/zenodo.6362349

♻ ☆ MutexMatch: Semi-Supervised Learning with Mutex-Based Consistency Regularization

Yue Duan, Zhen Zhao, Lei Qi, Lei Wang, Luping Zhou, Yinghuan Shi, Yang Gao

The core issue in semi-supervised learning (SSL) lies in how to effectively leverage unlabeled data, whereas most existing methods tend to put a great emphasis on the utilization of high-confidence samples yet seldom fully explore the usage of low-confidence samples. In this paper, we aim to utilize low-confidence samples in a novel way with our proposed mutex-based consistency regularization, namely MutexMatch. Specifically, the high-confidence samples are required to exactly predict "what it is" by conventional True-Positive Classifier, while the low-confidence samples are employed to achieve a simpler goal -- to predict with ease "what it is not" by True-Negative Classifier. In this sense, we not only mitigate the pseudo-labeling errors but also make full use of the low-confidence unlabeled data by consistency of dissimilarity degree. MutexMatch achieves superior performance on multiple benchmark datasets, i.e., CIFAR-10, CIFAR-100, SVHN, STL-10, mini-ImageNet and Tiny-ImageNet. More importantly, our method further shows superiority when the amount of labeled data is scarce, e.g., 92.23% accuracy with only 20 labeled data on CIFAR-10. Our code and model weights have been released at https://github.com/NJUyued/MutexMatch4SSL.

comment: Accepted by IEEE Transactions on Neural Networks and Learning Systems (TNNLS)

♻ ☆ Prototypical VoteNet for Few-Shot 3D Point Cloud Object Detection NeurIPS 2022

Shizhen Zhao, Xiaojuan Qi

Most existing 3D point cloud object detection approaches heavily rely on large amounts of labeled training data. However, the labeling process is costly and time-consuming. This paper considers few-shot 3D point cloud object detection, where only a few annotated samples of novel classes are needed with abundant samples of base classes. To this end, we propose Prototypical VoteNet to recognize and localize novel instances, which incorporates two new modules: Prototypical Vote Module (PVM) and Prototypical Head Module (PHM). Specifically, as the 3D basic geometric structures can be shared among categories, PVM is designed to leverage class-agnostic geometric prototypes, which are learned from base classes, to refine local features of novel categories.Then PHM is proposed to utilize class prototypes to enhance the global feature of each object, facilitating subsequent object localization and classification, which is trained by the episodic training strategy. To evaluate the model in this new setting, we contribute two new benchmark datasets, FS-ScanNet and FS-SUNRGBD. We conduct extensive experiments to demonstrate the effectiveness of Prototypical VoteNet, and our proposed method shows significant and consistent improvements compared to baselines on two benchmark datasets.

comment: NeurIPS 2022

♻ ☆ ADAS: A Simple Active-and-Adaptive Baseline for Cross-Domain 3D Semantic Segmentation

Ben Fei, Siyuan Huang, Jiakang Yuan, Botian Shi, Bo Zhang, Tao Chen, Min Dou, Yu Qiao

State-of-the-art 3D semantic segmentation models are trained on the off-the-shelf public benchmarks, but they often face the major challenge when these well-trained models are deployed to a new domain. In this paper, we propose an Active-and-Adaptive Segmentation (ADAS) baseline to enhance the weak cross-domain generalization ability of a well-trained 3D segmentation model, and bridge the point distribution gap between domains. Specifically, before the cross-domain adaptation stage begins, ADAS performs an active sampling operation to select a maximally-informative subset from both source and target domains for effective adaptation, reducing the adaptation difficulty under 3D scenarios. Benefiting from the rise of multi-modal 2D-3D datasets, ADAS utilizes a cross-modal attention-based feature fusion module that can extract a representative pair of image features and point features to achieve a bi-directional image-point feature interaction for better safe adaptation. Experimentally, ADAS is verified to be effective in many cross-domain settings including: 1) Unsupervised Domain Adaptation (UDA), which means that all samples from target domain are unlabeled; 2) Unsupervised Few-shot Domain Adaptation (UFDA) which means that only a few unlabeled samples are available in the unlabeled target domain; 3) Active Domain Adaptation (ADA) which means that the selected target samples by ADAS are manually annotated. Their results demonstrate that ADAS achieves a significant accuracy gain by easily coupling ADAS with self-training methods or off-the-shelf UDA works.

♻ ☆ Learning Generalized Visual Odometry Using Position-Aware Optical Flow and Geometric Bundle Adjustment

Yijun Cao, Xianshi Zhang, Fuya Luo, Peng Peng, Yongjie Li

Recent visual odometry (VO) methods incorporating geometric algorithm into deep-learning architecture have shown outstanding performance on the challenging monocular VO task. Despite encouraging results are shown, previous methods ignore the requirement of generalization capability under noisy environment and various scenes. To address this challenging issue, this work first proposes a novel optical flow network (PANet). Compared with previous methods that predict optical flow as a direct regression task, our PANet computes optical flow by predicting it into the discrete position space with optical flow probability volume, and then converting it to optical flow. Next, we improve the bundle adjustment module to fit the self-supervised training pipeline by introducing multiple sampling, ego-motion initialization, dynamic damping factor adjustment, and Jacobi matrix weighting. In addition, a novel normalized photometric loss function is advanced to improve the depth estimation accuracy. The experiments show that the proposed system not only achieves comparable performance with other state-of-the-art self-supervised learning-based methods on the KITTI dataset, but also significantly improves the generalization capability compared with geometry-based, learning-based and hybrid VO systems on the noisy KITTI and the challenging outdoor (KAIST) scenes.

comment: 35 pages, 6 figures

♻ ☆ Time to augment self-supervised visual representation learning

Arthur Aubret, Markus Ernst, Céline Teulière, Jochen Triesch

Biological vision systems are unparalleled in their ability to learn visual representations without supervision. In machine learning, self-supervised learning (SSL) has led to major advances in forming object representations in an unsupervised fashion. Such systems learn representations invariant to augmentation operations over images, like cropping or flipping. In contrast, biological vision systems exploit the temporal structure of the visual experience during natural interactions with objects. This gives access to "augmentations" not commonly used in SSL, like watching the same object from multiple viewpoints or against different backgrounds. Here, we systematically investigate and compare the potential benefits of such time-based augmentations during natural interactions for learning object categories. Our results show that time-based augmentations achieve large performance gains over state-of-the-art image augmentations. Specifically, our analyses reveal that: 1) 3-D object manipulations drastically improve the learning of object categories; 2) viewing objects against changing backgrounds is important for learning to discard background-related information from the latent representation. Overall, we conclude that time-based augmentations during natural interactions with objects can substantially improve self-supervised learning, narrowing the gap between artificial and biological vision systems.

comment: 20 pages

♻ ☆ Deep Generative Models on 3D Representations: A Survey

Zifan Shi, Sida Peng, Yinghao Xu, Yiyi Liao, Yujun Shen

Generative models, as an important family of statistical modeling, target learning the observed data distribution via generating new instances. Along with the rise of neural networks, deep generative models, such as variational autoencoders (VAEs) and generative adversarial network (GANs), have made tremendous progress in 2D image synthesis. Recently, researchers switch their attentions from the 2D space to the 3D space considering that 3D data better aligns with our physical world and hence enjoys great potential in practice. However, unlike a 2D image, which owns an efficient representation (i.e., pixel grid) by nature, representing 3D data could face far more challenges. Concretely, we would expect an ideal 3D representation to be capable enough to model shapes and appearances in details, and to be highly efficient so as to model high-resolution data with fast speed and low memory cost. However, existing 3D representations, such as point clouds, meshes, and recent neural fields, usually fail to meet the above requirements simultaneously. In this survey, we make a thorough review of the development of 3D generation, including 3D shape generation and 3D-aware image synthesis, from the perspectives of both algorithms and more importantly representations. We hope that our discussion could help the community track the evolution of this field and further spark some innovative ideas to advance this challenging task.

comment: Github: https://github.com/justimyhxu/awesome-3D-generation

♻ ☆ EZNAS: Evolving Zero Cost Proxies For Neural Architecture Scoring

Yash Akhauri, J. Pablo Munoz, Nilesh Jain, Ravi Iyer

Neural Architecture Search (NAS) has significantly improved productivity in the design and deployment of neural networks (NN). As NAS typically evaluates multiple models by training them partially or completely, the improved productivity comes at the cost of significant carbon footprint. To alleviate this expensive training routine, zero-shot/cost proxies analyze an NN at initialization to generate a score, which correlates highly with its true accuracy. Zero-cost proxies are currently designed by experts conducting multiple cycles of empirical testing on possible algorithms, datasets, and neural architecture design spaces. This experimentation lowers productivity and is an unsustainable approach towards zero-cost proxy design as deep learning use-cases diversify in nature. Additionally, existing zero-cost proxies fail to generalize across neural architecture design spaces. In this paper, we propose a genetic programming framework to automate the discovery of zero-cost proxies for neural architecture scoring. Our methodology efficiently discovers an interpretable and generalizable zero-cost proxy that gives state of the art score-accuracy correlation on all datasets and search spaces of NASBench-201 and Network Design Spaces (NDS). We believe that this research indicates a promising direction towards automatically discovering zero-cost proxies that can work across network architecture design spaces, datasets, and tasks.

♻ ☆ HouseCat6D -- A Large-Scale Multi-Modal Category Level 6D Object Pose Dataset with Household Objects in Realistic Scenarios

HyunJun Jung, Shun-Cheng Wu, Patrick Ruhkamp, Hannah Schieber, Pengyuan Wang, Giulia Rizzoli, Hongcheng Zhao, Sven Damian Meier, Daniel Roth, Nassir Navab, Benjamin Busam

Estimating the 6D pose of objects is one of the major fields in 3D computer vision. Since the promising outcomes from instance-level pose estimation, the research trends are heading towards category-level pose estimation for more practical application scenarios. However, unlike well-established instance-level pose datasets, available category-level datasets lack annotation quality and provided pose quantity. We propose the new category level 6D pose dataset HouseCat6D featuring 1) Multi-modality of Polarimetric RGB+P and Depth, 2) Highly diverse 194 objects of 10 household object categories including 2 photometrically challenging categories, 3) High-quality pose annotation with an error range of only 1.35 mm to 1.74 mm, 4) 41 large scale scenes with extensive viewpoint coverage, 5) Checkerboard-free environment throughout the entire scene. We also provide benchmark results of state-of-the-art category-level pose estimation networks.

♻ ☆ Improving video retrieval using multilingual knowledge transfer

Avinash Madasu, Estelle Aflalo, Gabriela Ben Melech Stan, Shao-Yen Tseng, Gedas Bertasius, Vasudev Lal

Video retrieval has seen tremendous progress with the development of vision-language models. However, further improving these models require additional labelled data which is a huge manual effort. In this paper, we propose a framework MKTVR, that utilizes knowledge transfer from a multilingual model to boost the performance of video retrieval. We first use state-of-the-art machine translation models to construct pseudo ground-truth multilingual video-text pairs. We then use this data to learn a video-text representation where English and non-English text queries are represented in a common embedding space based on pretrained multilingual models. We evaluate our proposed approach on four English video retrieval datasets such as MSRVTT, MSVD, DiDeMo and Charades. Experimental results demonstrate that our approach achieves state-of-the-art results on all datasets outperforming previous models. Finally, we also evaluate our model on a multilingual video-retrieval dataset encompassing six languages and show that our model outperforms previous multilingual video retrieval models in a zero-shot setting.

♻ ☆ Recent Advances in Scene Image Representation and Classification

Chiranjibi Sitaula, Tej Bahadur Shahi, Faezeh Marzbanrad, Jagannath Aryal

With the rise of deep learning algorithms nowadays, scene image representation methods have achieved a significant performance boost in classification. However, the performance is still limited because the scene images are mostly complex having higher intra-class dissimilarity and inter-class similarity problems. To deal with such problems, there have been several methods proposed in the literature with their advantages and limitations. A detailed study of previous works is necessary to understand their advantages and disadvantages in image representation and classification problems. In this paper, we review the existing scene image representation methods that are being widely used for image classification. For this, we, first, devise the taxonomy using the seminal existing methods proposed in the literature to this date {using deep learning (DL)-based, computer vision (CV)-based, and search engine (SE)-based methods}. Next, we compare their performance both qualitatively (e.g., quality of outputs, pros/cons, etc.) and quantitatively (e.g., accuracy). Last, we speculate on the prominent research directions in scene image representation tasks using {keyword growth and timeline analysis.} Overall, this survey provides in-depth insights and applications of recent scene image representation methods under three different methods.

comment: This paper is under review in Multimedia Tools and Applications (Springer) journal. This article may be deleted or updated based on the policies of the journal

♻ ☆ A Comprehensive Survey of Data Augmentation in Visual Reinforcement Learning

Guozheng Ma, Zhen Wang, Zhecheng Yuan, Xueqian Wang, Bo Yuan, Dacheng Tao

Visual reinforcement learning (RL), which makes decisions directly from high-dimensional visual inputs, has demonstrated significant potential in various domains. However, deploying visual RL techniques in the real world remains challenging due to their low sample efficiency and large generalization gaps. To tackle these obstacles, data augmentation (DA) has become a widely used technique in visual RL for acquiring sample-efficient and generalizable policies by diversifying the training data. This survey aims to provide a timely and essential review of DA techniques in visual RL in recognition of the thriving development in this field. In particular, we propose a unified framework for analyzing visual RL and understanding the role of DA in it. We then present a principled taxonomy of the existing augmentation techniques used in visual RL and conduct an in-depth discussion on how to better leverage augmented data in different scenarios. Moreover, we report a systematic empirical evaluation of DA-based techniques in visual RL and conclude by highlighting the directions for future research. As the first comprehensive survey of DA in visual RL, this work is expected to offer valuable guidance to this emerging field.

comment: A well-classified paper list that will be continuously updated can be found at https://github.com/Guozheng-Ma/DA-in-visualRL

♻ ☆ ColoristaNet for Photorealistic Video Style Transfer

Xiaowen Qiu, Ruize Xu, Boan He, Yingtao Zhang, Wenqiang Zhang, Weifeng Ge

Photorealistic style transfer aims to transfer the artistic style of an image onto an input image or video while keeping photorealism. In this paper, we think it's the summary statistics matching scheme in existing algorithms that leads to unrealistic stylization. To avoid employing the popular Gram loss, we propose a self-supervised style transfer framework, which contains a style removal part and a style restoration part. The style removal network removes the original image styles, and the style restoration network recovers image styles in a supervised manner. Meanwhile, to address the problems in current feature transformation methods, we propose decoupled instance normalization to decompose feature transformation into style whitening and restylization. It works quite well in ColoristaNet and can transfer image styles efficiently while keeping photorealism. To ensure temporal coherency, we also incorporate optical flow methods and ConvLSTM to embed contextual information. Experiments demonstrates that ColoristaNet can achieve better stylization effects when compared with state-of-the-art algorithms.

comment: 30 pages, 29 figures

♻ ☆ Pho(SC)-CTC -- A Hybrid Approach Towards Zero-shot Word Image Recognition SC

Ravi Bhatt, Anuj Rai, Narayanan C. Krishnan, Sukalpa Chanda

Annotating words in a historical document image archive for word image recognition purpose demands time and skilled human resource (like historians, paleographers). In a real-life scenario, obtaining sample images for all possible words is also not feasible. However, Zero-shot learning methods could aptly be used to recognize unseen/out-of-lexicon words in such historical document images. Based on previous state-of-the-art method for zero-shot word recognition Pho(SC)Net, we propose a hybrid model based on the CTC framework (Pho(SC)-CTC) that takes advantage of the rich features learned by Pho(SC)Net followed by a connectionist temporal classification (CTC) framework to perform the final classification. Encouraging results were obtained on two publicly available historical document datasets and one synthetic handwritten dataset, which justifies the efficacy of Pho(SC)-CTC and Pho(SC)Net.

comment: Accepted (International Journal on Document Analysis and Recognition). This paper is the extension of the paper titled "Pho(SC)Net: An Approach Towards Zero-shot Word Image Recognition in Historical Documents" published in ICDAR 2021

♻ ☆ MMDialog: A Large-scale Multi-turn Dialogue Dataset Towards Multi-modal Open-domain Conversation

Jiazhan Feng, Qingfeng Sun, Can Xu, Pu Zhao, Yaming Yang, Chongyang Tao, Dongyan Zhao, Qingwei Lin

Responding with multi-modal content has been recognized as an essential capability for an intelligent conversational agent. In this paper, we introduce the MMDialog dataset to better facilitate multi-modal conversation. MMDialog is composed of a curated set of 1.08 million real-world dialogues with 1.53 million unique images across 4,184 topics. MMDialog has two main and unique advantages. First, it is the largest multi-modal conversation dataset by the number of dialogues by 88x. Second, it contains massive topics to generalize the open-domain. To build engaging dialogue system with this dataset, we propose and normalize two response producing tasks based on retrieval and generative scenarios. In addition, we build two baselines for above tasks with state-of-the-art techniques and report their experimental performance. We also propose a novel evaluation metric MM-Relevance to measure the multi-modal responses. Our dataset and scripts are available in https://github.com/victorsungo/MMDialog.

♻ ☆ MGTANet: Encoding Sequential LiDAR Points Using Long Short-Term Motion-Guided Temporal Attention for 3D Object Detection AAAI

Junho Koh, Junhyung Lee, Youngwoo Lee, Jaekyum Kim, Jun Won Choi

Most scanning LiDAR sensors generate a sequence of point clouds in real-time. While conventional 3D object detectors use a set of unordered LiDAR points acquired over a fixed time interval, recent studies have revealed that substantial performance improvement can be achieved by exploiting the spatio-temporal context present in a sequence of LiDAR point sets. In this paper, we propose a novel 3D object detection architecture, which can encode LiDAR point cloud sequences acquired by multiple successive scans. The encoding process of the point cloud sequence is performed on two different time scales. We first design a short-term motion-aware voxel encoding that captures the short-term temporal changes of point clouds driven by the motion of objects in each voxel. We also propose long-term motion-guided bird's eye view (BEV) feature enhancement that adaptively aligns and aggregates the BEV feature maps obtained by the short-term voxel encoding by utilizing the dynamic motion context inferred from the sequence of the feature maps. The experiments conducted on the public nuScenes benchmark demonstrate that the proposed 3D object detector offers significant improvements in performance compared to the baseline methods and that it sets a state-of-the-art performance for certain 3D object detection categories. Code is available at https://github.com/HYjhkoh/MGTANet.git

comment: Thirty-Seventh AAAI Conference on Artificial Intelligence (AAAI'23)

♻ ☆ Multi-branch Cascaded Swin Transformers with Attention to k-space Sampling Pattern for Accelerated MRI Reconstruction

Mevan Ekanayake, Kamlesh Pawar, Mehrtash Harandi, Gary Egan, Zhaolin Chen

Global correlations are widely seen in human anatomical structures due to similarity across tissues and bones. These correlations are reflected in magnetic resonance imaging (MRI) scans as a result of close-range proton density and T1/T2 parameters. Furthermore, to achieve accelerated MRI, k-space data are undersampled which causes global aliasing artifacts. Convolutional neural network (CNN) models are widely utilized for accelerated MRI reconstruction, but those models are limited in capturing global correlations due to the intrinsic locality of the convolution operation. The self-attention-based transformer models are capable of capturing global correlations among image features, however, the current contributions of transformer models for MRI reconstruction are minute. The existing contributions mostly provide CNN-transformer hybrid solutions and rarely leverage the physics of MRI. In this paper, we propose a physics-based stand-alone (convolution free) transformer model titled, the Multi-head Cascaded Swin Transformers (McSTRA) for accelerated MRI reconstruction. McSTRA combines several interconnected MRI physics-related concepts with the transformer networks: it exploits global MR features via the shifted window self-attention mechanism; it extracts MR features belonging to different spectral components separately using a multi-head setup; it iterates between intermediate de-aliasing and k-space correction via a cascaded network with data consistency in k-space and intermediate loss computations; furthermore, we propose a novel positional embedding generation mechanism to guide self-attention utilizing the point spread function corresponding to the undersampling mask. Our model significantly outperforms state-of-the-art MRI reconstruction methods both visually and quantitatively while depicting improved resolution and removal of aliasing artifacts.

♻ ☆ Learning Temporally and Semantically Consistent Unpaired Video-to-video Translation Through Pseudo-Supervision From Synthetic Optical Flow

Kaihong Wang, Kumar Akash, Teruhisa Misu

Unpaired video-to-video translation aims to translate videos between a source and a target domain without the need of paired training data, making it more feasible for real applications. Unfortunately, the translated videos generally suffer from temporal and semantic inconsistency. To address this, many existing works adopt spatiotemporal consistency constraints incorporating temporal information based on motion estimation. However, the inaccuracies in the estimation of motion deteriorate the quality of the guidance towards spatiotemporal consistency, which leads to unstable translation. In this work, we propose a novel paradigm that regularizes the spatiotemporal consistency by synthesizing motions in input videos with the generated optical flow instead of estimating them. Therefore, the synthetic motion can be applied in the regularization paradigm to keep motions consistent across domains without the risk of errors in motion estimation. Thereafter, we utilize our unsupervised recycle and unsupervised spatial loss, guided by the pseudo-supervision provided by the synthetic optical flow, to accurately enforce spatiotemporal consistency in both domains. Experiments show that our method is versatile in various scenarios and achieves state-of-the-art performance in generating temporally and semantically consistent videos. Code is available at: https://github.com/wangkaihong/Unsup_Recycle_GAN/.

♻ ☆ Exploring Consistency in Cross-Domain Transformer for Domain Adaptive Semantic Segmentation

Kaihong Wang, Donghyun Kim, Rogerio Feris, Kate Saenko, Margrit Betke

While transformers have greatly boosted performance in semantic segmentation, domain adaptive transformers are not yet well explored. We identify that the domain gap can cause discrepancies in self-attention. Due to this gap, the transformer attends to spurious regions or pixels, which deteriorates accuracy on the target domain. We propose to perform adaptation on attention maps with cross-domain attention layers that share features between the source and the target domains. Specifically, we impose consistency between predictions from cross-domain attention and self-attention modules to encourage similar distribution in the attention and output of the model across domains, i.e., attention-level and output-level alignment. We also enforce consistency in attention maps between different augmented views to further strengthen the attention-based alignment. Combining these two components, our method mitigates the discrepancy in attention maps across domains and further boosts the performance of the transformer under unsupervised domain adaptation settings. Our model outperforms the existing state-of-the-art baseline model on three widely used benchmarks, including GTAV-to-Cityscapes by 1.3 percent point (pp), Synthia-to-Cityscapes by 0.6 pp, and Cityscapes-to-ACDC by 1.1 pp, on average. Additionally, we verify the effectiveness and generalizability of our method through extensive experiments. Our code will be publicly available.

♻ ☆ Content Adaptive Latents and Decoder for Neural Image Compression ECCV 2022

Guanbo Pan, Guo Lu, Zhihao Hu, Dong Xu

In recent years, neural image compression (NIC) algorithms have shown powerful coding performance. However, most of them are not adaptive to the image content. Although several content adaptive methods have been proposed by updating the encoder-side components, the adaptability of both latents and the decoder is not well exploited. In this work, we propose a new NIC framework that improves the content adaptability on both latents and the decoder. Specifically, to remove redundancy in the latents, our content adaptive channel dropping (CACD) method automatically selects the optimal quality levels for the latents spatially and drops the redundant channels. Additionally, we propose the content adaptive feature transformation (CAFT) method to improve decoder-side content adaptability by extracting the characteristic information of the image content, which is then used to transform the features in the decoder side. Experimental results demonstrate that our proposed methods with the encoder-side updating algorithm achieve the state-of-the-art performance.

comment: V1 is accepted to ECCV 2022. V2 is the improved version

♻ ☆ Synthesizing Informative Training Samples with GAN NeurIPS 2022

Bo Zhao, Hakan Bilen

Remarkable progress has been achieved in synthesizing photo-realistic images with generative adversarial networks (GANs). Recently, GANs are utilized as the training sample generator when obtaining or storing real training data is expensive even infeasible. However, traditional GANs generated images are not as informative as the real training samples when being used to train deep neural networks. In this paper, we propose a novel method to synthesize Informative Training samples with GAN (IT-GAN). Specifically, we freeze a pre-trained GAN model and learn the informative latent vectors that correspond to informative training samples. The synthesized images are required to preserve information for training deep neural networks rather than visual reality or fidelity. Experiments verify that the deep neural networks can learn faster and achieve better performance when being trained with our IT-GAN generated images. We also show that our method is a promising solution to dataset condensation problem.

comment: NeurIPS 2022 Workshop on Synthetic Data for Empowering ML Research, https://openreview.net/forum?id=frAv0jtUMfS

♻ ☆ Simple Neighborhood Representative Pre-processing Boosts Outlier Detectors

Jiawei Yang, Yu Chen, Sylwan Rahardja

Over the decades, traditional outlier detectors have ignored the group-level factor when calculating outlier scores for objects in data by evaluating only the object-level factor, failing to capture the collective outliers. To mitigate this issue, we present a method called neighborhood representative (NR), which empowers all the existing outlier detectors to efficiently detect outliers, including collective outliers, while maintaining their computational integrity. It achieves this by selecting representative objects, scoring these objects, then applies the score of the representative objects to its collective objects. Without altering existing detectors, NR is compatible with existing detectors, while improving performance on real world datasets with +8% (0.72 to 0.78 AUC) relative to state-of-the-art outlier detectors.

♻ ☆ Generalized Object Detection on Fisheye Cameras for Autonomous Driving: Dataset, Representations and Baseline

Hazem Rashed, Eslam Mohamed, Ganesh Sistu, Varun Ravi Kumar, Ciaran Eising, Ahmad El-Sallab, Senthil Yogamani

Object detection is a comprehensively studied problem in autonomous driving. However, it has been relatively less explored in the case of fisheye cameras. The standard bounding box fails in fisheye cameras due to the strong radial distortion, particularly in the image's periphery. We explore better representations like oriented bounding box, ellipse, and generic polygon for object detection in fisheye images in this work. We use the IoU metric to compare these representations using accurate instance segmentation ground truth. We design a novel curved bounding box model that has optimal properties for fisheye distortion models. We also design a curvature adaptive perimeter sampling method for obtaining polygon vertices, improving relative mAP score by 4.9% compared to uniform sampling. Overall, the proposed polygon model improves mIoU relative accuracy by 40.3%. It is the first detailed study on object detection on fisheye cameras for autonomous driving scenarios to the best of our knowledge. The dataset comprising of 10,000 images along with all the object representations ground truth will be made public to encourage further research. We summarize our work in a short video with qualitative results at https://youtu.be/iLkOzvJpL-A.

comment: Camera ready version. Accepted for presentation at Winter Conference on Applications of Computer Vision 2021. Dataset is shared at https://drive.google.com/drive/folders/1bobmY2wlIBozeU5ZgPfYPqVAnpPw4QrM

♻ ☆ Audio-Visual Person-of-Interest DeepFake Detection

Davide Cozzolino, Matthias Nießner, Luisa Verdoliva

Face manipulation technology is advancing very rapidly, and new methods are being proposed day by day. The aim of this work is to propose a deepfake detector that can cope with the wide variety of manipulation methods and scenarios encountered in the real world. Our key insight is that each person has specific biometric characteristics that a synthetic generator cannot likely reproduce. Accordingly, we extract high-level audio-visual biometric features which characterize the identity of a person, and use them to create a person-of-interest (POI) deepfake detector. We leverage a contrastive learning paradigm to learn the moving-face and audio segment embeddings that are most discriminative for each identity. As a result, when the video and/or audio of a person is manipulated, its representation in the embedding space becomes inconsistent with the real identity, allowing reliable detection. Training is carried out exclusively on real talking-face videos, thus the detector does not depend on any specific manipulation method and yields the highest generalization ability. In addition, our method can detect both single-modality (audio-only, video-only) and multi-modality (audio-video) attacks, and is robust to low-quality or corrupted videos by building only on high-level semantic features. Experiments on a wide variety of datasets confirm that our method ensures a SOTA performance, with an average improvement in terms of AUC of around 3%, 10%, and 4% for high-quality, low quality, and attacked videos, respectively. https://github.com/grip-unina/poi-forensics

♻ ☆ An Online Learning System for Wireless Charging Alignment using Surround-view Fisheye Cameras

Ashok Dahal, Varun Ravi Kumar, Senthil Yogamani, Ciaran Eising

Electric Vehicles are increasingly common, with inductive chargepads being considered a convenient and efficient means of charging electric vehicles. However, drivers are typically poor at aligning the vehicle to the necessary accuracy for efficient inductive charging, making the automated alignment of the two charging plates desirable. In parallel to the electrification of the vehicular fleet, automated parking systems that make use of surround-view camera systems are becoming increasingly popular. In this work, we propose a system based on the surround-view camera architecture to detect, localize, and automatically align the vehicle with the inductive chargepad. The visual design of the chargepads is not standardized and not necessarily known beforehand. Therefore, a system that relies on offline training will fail in some situations. Thus, we propose a self-supervised online learning method that leverages the driver's actions when manually aligning the vehicle with the chargepad and combine it with weak supervision from semantic segmentation and depth to learn a classifier to auto-annotate the chargepad in the video for further training. In this way, when faced with a previously unseen chargepad, the driver needs only manually align the vehicle a single time. As the chargepad is flat on the ground, it is not easy to detect it from a distance. Thus, we propose using a Visual SLAM pipeline to learn landmarks relative to the chargepad to enable alignment from a greater range. We demonstrate the working system on an automated vehicle as illustrated in the video at https://youtu.be/_cLCmkW4UYo. To encourage further research, we will share a chargepad dataset used in this work.

comment: Accepted for publication at IEEE Transactions on Intelligent Transportation Systems. Chargepad Dataset is shared at https://drive.google.com/drive/folders/1KeLFIqOnhU2CGsD0vbiN9UqKmBSyHERd

♻ ☆ A Multiple-View Geometric Model for Specularity Prediction on General Curved Surfaces

Alexandre Morgand, Mohamed Tamaazousti, Adrien Bartoli

Specularity prediction is essential to many computer vision applications, giving important visual cues usable in Augmented Reality (AR), Simultaneous Localisation and Mapping (SLAM), 3D reconstruction and material modeling. However, it is a challenging task requiring numerous information from the scene including the camera pose, the geometry of the scene, the light sources and the material properties. Our previous work addressed this task by creating an explicit model using an ellipsoid whose projection fits the specularity image contours for a given camera pose. These ellipsoid-based approaches belong to a family of models called JOint-LIght MAterial Specularity (JOLIMAS), which we have gradually improved by removing assumptions on the scene geometry. However, our most recent approach is still limited to uniformly curved surfaces. This paper generalises JOLIMAS to any surface geometry while improving the quality of specularity prediction, without sacrificing computation performances. The proposed method establishes a link between surface curvature and specularity shape in order to lift the geometric assumptions made in previous work. Contrary to previous work, our new model is built from a physics-based local illumination model namely Torrance-Sparrow, providing an improved reconstruction. Specularity prediction using our new model is tested against the most recent JOLIMAS version on both synthetic and real sequences with objects of various general shapes. Our method outperforms previous approaches in specularity prediction, including the real-time setup, as shown in the supplementary videos.

♻ ☆ Query Semantic Reconstruction for Background in Few-Shot Segmentation

Haoyan Guan, Michael Spratling

Few-shot segmentation (FSS) aims to segment unseen classes using a few annotated samples. Typically, a prototype representing the foreground class is extracted from annotated support image(s) and is matched to features representing each pixel in the query image. However, models learnt in this way are insufficiently discriminatory, and often produce false positives: misclassifying background pixels as foreground. Some FSS methods try to address this issue by using the background in the support image(s) to help identify the background in the query image. However, the backgrounds of theses images is often quite distinct, and hence, the support image background information is uninformative. This article proposes a method, QSR, that extracts the background from the query image itself, and as a result is better able to discriminate between foreground and background features in the query image. This is achieved by modifying the training process to associate prototypes with class labels including known classes from the training data and latent classes representing unknown background objects. This class information is then used to extract a background prototype from the query image. To successfully associate prototypes with class labels and extract a background prototype that is capable of predicting a mask for the background regions of the image, the machinery for extracting and using foreground prototypes is induced to become more discriminative between different classes. Experiments for both 1-shot and 5-shot FSS on both the PASCAL-5i and COCO-20i datasets demonstrate that the proposed method results in a significant improvement in performance for the baseline methods it is applied to. As QSR operates only during training, these improved results are produced with no extra computational complexity during testing.

☆ Is it worth it? An experimental comparison of six deep- and classical machine learning methods for unsupervised anomaly detection in time series

Ferdinand Rewicki, Joachim Denzler, Julia Niebling

The detection of anomalies in time series data is crucial in a wide range of applications, such as system monitoring, health care or cyber security. While the vast number of available methods makes selecting the right method for a certain application hard enough, different methods have different strengths, e.g. regarding the type of anomalies they are able to find. In this work, we compare six unsupervised anomaly detection methods with different complexities to answer the questions: Are the more complex methods usually performing better? And are there specific anomaly types that those method are tailored to? The comparison is done on the UCR anomaly archive, a recent benchmark dataset for anomaly detection. We compare the six methods by analyzing the experimental results on a dataset- and anomaly type level after tuning the necessary hyperparameter for each method. Additionally we examine the ability of individual methods to incorporate prior knowledge about the anomalies and analyse the differences of point-wise and sequence wise features. We show with broad experiments, that the classical machine learning methods show a superior performance compared to the deep learning methods across a wide range of anomaly types.

comment: 15 Pages, The repository to reproduce the results is available at https://gitlab.com/dlr-dw/is-it-worth-it-benchmark

☆ The Ties that matter: From the perspective of Similarity Measure in Online Social Networks

Soumita Das, Anupam Biswas

Online Social Networks have embarked on the importance of connection strength measures which has a broad array of applications such as, analyzing diffusion behaviors, community detection, link predictions, recommender systems. Though there are some existing connection strength measures, the density that a connection shares with it's neighbors and the directionality aspect has not received much attention. In this paper, we have proposed an asymmetric edge similarity measure namely, Neighborhood Density-based Edge Similarity (NDES) which provides a fundamental support to derive the strength of connection. The time complexity of NDES is $O(nk^2)$. An application of NDES for community detection in social network is shown. We have considered a similarity based community detection technique and substituted its similarity measure with NDES. The performance of NDES is evaluated on several small real-world datasets in terms of the effectiveness in detecting communities and compared with three widely used similarity measures. Empirical results show NDES enables detecting comparatively better communities both in terms of accuracy and quality.

comment: To be published in MINDS-2021

☆ DCC: A Cascade based Approach to Detect Communities in Social Networks

Soumita Das, Anupam Biswas, Akrati Saxena

Community detection in Social Networks is associated with finding and grouping the most similar nodes inherent in the network. These similar nodes are identified by computing tie strength. Stronger ties indicates higher proximity shared by connected node pairs. This work is motivated by Granovetter's argument that suggests that strong ties lies within densely connected nodes and the theory that community cores in real-world networks are densely connected. In this paper, we have introduced a novel method called \emph{Disjoint Community detection using Cascades (DCC)} which demonstrates the effectiveness of a new local density based tie strength measure on detecting communities. Here, tie strength is utilized to decide the paths followed for propagating information. The idea is to crawl through the tuple information of cascades towards the community core guided by increasing tie strength. Considering the cascade generation step, a novel preferential membership method has been developed to assign community labels to unassigned nodes. The efficacy of $DCC$ has been analyzed based on quality and accuracy on several real-world datasets and baseline community detection algorithms.

comment: To be published in CHSN-2022

☆ RECAP: Retrieval Augmented Music Captioner

Zihao He, Weituo Hao, Xuchen Song

With the prevalence of stream media platforms serving music search and recommendation, interpreting music by understanding audio and lyrics interactively has become an important and challenging task. However, many previous works focus on refining individual components of encoder-decoder architecture mapping music to caption tokens, ignoring the potential usage of audio and lyrics correspondence. In this paper, we propose to explicitly learn the multi-modal alignment with retrieval augmentation by contrastive learning. By learning audio-lyrics correspondence, the model is guided to learn better cross-modal attention weights, thus generating high-quality caption words. We provide both theoretical and empirical results that demonstrate the advantage of the proposed method.

☆ Beyond Information Exchange: An Approach to Deploy Network Properties for Information Diffusion

Soumita Das, Anupam Biswas, Ravi Kishore Devarapalli

Information diffusion in Online Social Networks is a new and crucial problem in social network analysis field and requires significant research attention. Efficient diffusion of information are of critical importance in diverse situations such as; pandemic prevention, advertising, marketing etc. Although several mathematical models have been developed till date, but previous works lacked systematic analysis and exploration of the influence of neighborhood for information diffusion. In this paper, we have proposed Common Neighborhood Strategy (CNS) algorithm for information diffusion that demonstrates the role of common neighborhood in information propagation throughout the network. The performance of CNS algorithm is evaluated on several real-world datasets in terms of diffusion speed and diffusion outspread and compared with several widely used information diffusion models. Empirical results show CNS algorithm enables better information diffusion both in terms of diffusion speed and diffusion outspread.

comment: To be published in BigDML 2021

☆ Multi-hop Evidence Retrieval for Cross-document Relation Extraction

Keming Lu, I-Hung Hsu, Wenxuan Zhou, Mingyu Derek Ma, Muhao Chen

Relation Extraction (RE) has been extended to cross-document scenarios because many relations are not simply described in a single document. This inevitably brings the challenge of efficient open-space evidence retrieval to support the inference of cross-document relations, along with the challenge of multi-hop reasoning on top of entities and evidence scattered in an open set of documents. To combat these challenges, we propose Mr.CoD, a multi-hop evidence retrieval method based on evidence path mining and ranking with adapted dense retrievers. We explore multiple variants of retrievers to show evidence retrieval is an essential part in cross-document RE. Experiments on CodRED show that evidence retrieval with Mr.Cod effectively acquires cross-document evidence that essentially supports open-setting cross-document RE. Additionally, we show that Mr.CoD facilitates evidence retrieval and boosts end-to-end RE performance with effective multi-hop reasoning in both closed and open settings of RE.

comment: Work in progress

☆ Learning List-Level Domain-Invariant Representations for Ranking

Ruicheng Xian, Honglei Zhuang, Zhen Qin, Hamed Zamani, Jing Lu, Ji Ma, Kai Hui, Han Zhao, Xuanhui Wang, Michael Bendersky

Domain adaptation aims to transfer the knowledge acquired by models trained on (data-rich) source domains to (low-resource) target domains, for which a popular method is invariant representation learning. While they have been studied extensively for classification and regression problems, how they apply to ranking problems, where the data and metrics have a list structure, is not well understood. Theoretically, we establish a domain adaptation generalization bound for ranking under listwise metrics such as MRR and NDCG. The bound suggests an adaptation method via learning list-level domain-invariant feature representations, whose benefits are empirically demonstrated by unsupervised domain adaptation experiments on real-world ranking tasks, including passage reranking. A key message is that for domain adaptation, the representations should be analyzed at the same level at which the metric is computed, as we show that learning invariant representations at the list level is most effective for adaptation on ranking problems.

☆ AgAsk: An Agent to Help Answer Farmer's Questions From Scientific Documents

Bevan Koopman, Ahmed Mourad, Hang Li, Anton van der Vegt, Shengyao Zhuang, Simon Gibson, Yash Dang, David Lawrence, Guido Zuccon

Decisions in agriculture are increasingly data-driven; however, valuable agricultural knowledge is often locked away in free-text reports, manuals and journal articles. Specialised search systems are needed that can mine agricultural information to provide relevant answers to users' questions. This paper presents AgAsk -- an agent able to answer natural language agriculture questions by mining scientific documents. We carefully survey and analyse farmers' information needs. On the basis of these needs we release an information retrieval test collection comprising real questions, a large collection of scientific documents split in passages, and ground truth relevance assessments indicating which passages are relevant to each question. We implement and evaluate a number of information retrieval models to answer farmers questions, including two state-of-the-art neural ranking models. We show that neural rankers are highly effective at matching passages to questions in this context. Finally, we propose a deployment architecture for AgAsk that includes a client based on the Telegram messaging platform and retrieval model deployed on commodity hardware. The test collection we provide is intended to stimulate more research in methods to match natural language to answers in scientific documents. While the retrieval models were evaluated in the agriculture domain, they are generalisable and of interest to others working on similar problems. The test collection is available at: \url{https://github.com/ielab/agvaluate}.

comment: 17 pages, submitted to IJDL

☆ Audio Denoising for Robust Audio Fingerprinting

Kamil Akesbi

Music discovery services let users identify songs from short mobile recordings. These solutions are often based on Audio Fingerprinting, and rely more specifically on the extraction of spectral peaks in order to be robust to a number of distortions. Few works have been done to study the robustness of these algorithms to background noise captured in real environments. In particular, AFP systems still struggle when the signal to noise ratio is low, i.e when the background noise is strong. In this project, we tackle this problematic with Deep Learning. We test a new hybrid strategy which consists of inserting a denoising DL model in front of a peak-based AFP algorithm. We simulate noisy music recordings using a realistic data augmentation pipeline, and train a DL model to denoise them. The denoising model limits the impact of background noise on the AFP system's extracted peaks, improving its robustness to noise. We further propose a novel loss function to adapt the DL model to the considered AFP system, increasing its precision in terms of retrieved spectral peaks. To the best of our knowledge, this hybrid strategy has not been tested before.

comment: 63 pages, master thesis

♻ ☆ COVID-19 Activity Risk Calculator as a Gamified Public Health Intervention Tool

Shreyasvi Natraj, Malhar Bhide, Nathan Yap, Meng Liu, Agrima Seth, Christin Glorioso

Public health intervention techniques have been highly significant in reducing the negative impact of several epidemics and pandemics. Among all of the wide-spread diseases, one of the most dangerous one has been severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) or Coronavirus disease 2019 (COVID-19). The impact of the virus has been observed in over 200 countries leading to hospitalizations and deaths of millions of people. Currently existing COVID-19 risk estimation tools provided to the general public have been highly variable during the pandemic due to its dependency on rapidly evolving factors such as community transmission levels and variants. There has also been confusion surrounding certain personal protective strategies such as risk reduction by mask-wearing and vaccination. In order to create a simplified easy-to-use tool for estimating different individual risks associated with carrying out daily-life activity, we developed COVID-19 Activity Risk Calculator (CovARC). CovARC serves as a gamified public health intervention as users can "play with" how different risks associated with COVID-19 would change depending on several different factors when carrying out a daily routine activity. Empowering the public to make informed, data-driven decisions about safely engaging in activities may help to reduce COVID- 19 levels in the community. In this study, we demonstrate a streamlined, scalable and accurate COVID-19 risk calculation system. Our study also showcases quantitatively, the increased impact of interventions such as vaccination and mask-wearing when cases are higher, which could prove as a validity to inform and support policy decisions around mask mandate case thresholds and other non-pharmaceutical interventions.

comment: 11 pages, 6 figures (main paper + 1 figure supplementary section.)

♻ ☆ Rank4Class: A Ranking Formulation for Multiclass Classification

Nan Wang, Zhen Qin, Le Yan, Honglei Zhuang, Xuanhui Wang, Michael Bendersky, Marc Najork

Multiclass classification (MCC) is a fundamental machine learning problem of classifying each instance into one of a predefined set of classes. In the deep learning era, extensive efforts have been spent on developing more powerful neural embedding models to better represent the instance for improving MCC performance. In this paper, we do not aim to propose new neural models for instance representation learning, but to show that it is promising to boost MCC performance with a novel formulation through the lens of ranking. In particular, by viewing MCC as to rank classes for an instance, we first argue that ranking metrics, such as Normalized Discounted Cumulative Gain, can be more informative than the commonly used Top-$K$ metrics. We further demonstrate that the dominant neural MCC recipe can be transformed to a neural ranking framework. Based on such generalization, we show that it is intuitive to leverage advanced techniques from the learning to rank literature to improve the MCC performance out of the box. Extensive empirical results on both text and image classification tasks with diverse datasets and backbone neural models show the value of our proposed framework.

♻ ☆ Data-Augmented Counterfactual Learning for Bundle Recommendation

Shixuan Zhu, Qi Shen, Yiming Zhang, Zhenwei Dong, Zhihua Wei

Bundle Recommendation (BR) aims at recommending bundled items on online content or e-commerce platform, such as song lists on a music platform or book lists on a reading website. Several graph based models have achieved state-of-the-art performance on BR task. But their performance is still sub-optimal, since the data sparsity problem tends to be more severe in real bundle recommendation scenarios, which limits graph-based models from more sufficient learning. In this paper, we propose a novel graph learning paradigm called Counterfactual Learning for Bundle Recommendation (CLBR) to mitigate the impact of data sparsity problem and improve bundle recommendation. Our paradigm consists of two main parts: counterfactual data augmentation and counterfactual constraint. The main idea of our paradigm lies in answering the counterfactual questions: "What would a user interact with if his/her interaction history changes?" "What would a user interact with if the bundle-item affiliation relations change?" In counterfactual data augmentation, we design a heuristic sampler to generate counterfactual graph views for graph-based models, which has better noise controlling than the stochastic sampler. We further propose counterfactual loss to constrain model learning for mitigating the effects of residual noise in augmented data and achieving more sufficient model optimization. Further theoretical analysis demonstrates the rationality of our design. Extensive experiments of BR models applied with our paradigm on two real-world datasets are conducted to verify the effectiveness of the paradigm.

comment: 9 pages

♻ ☆ Fairness in Recommender Systems: Research Landscape and Future Directions

Yashar Deldjoo, Dietmar Jannach, Alejandro Bellogin, Alessandro Difonzo, Dario Zanzonelli

Recommender systems can strongly influence which information we see online, e.g., on social media, and thus impact our beliefs, decisions, and actions. At the same time, these systems can create substantial business value for different stakeholders. Given the growing potential impact of such AI-based systems on individuals, organizations, and society, questions of fairness have gained increased attention in recent years. However, research on fairness in recommender systems is still a developing area. In this survey, we first review the fundamental concepts and notions of fairness that were put forward in the area in the recent past. Afterward, through a review of more than 150 scholarly publications, we present an overview of how research in this field is currently operationalized, e.g., in terms of general research methodology, fairness measures, and algorithmic approaches. Overall, our analysis of recent works points to specific research gaps. In particular, we find that in many research works in computer science, very abstract problem operationalizations are prevalent, and questions of the underlying normative claims and what represents a fair recommendation in the context of a given application are often not discussed in depth. These observations call for more interdisciplinary research to address fairness in recommendation in a more comprehensive and impactful manner.

♻ ☆ Optimizing Test-Time Query Representations for Dense Retrieval

Mujeen Sung, Jungsoo Park, Jaewoo Kang, Danqi Chen, Jinhyuk Lee

Recent developments of dense retrieval rely on quality representations of queries and contexts coming from pre-trained query and context encoders. In this paper, we introduce TouR (test-time optimization of query representations), which further optimizes instance-level query representations guided by signals from test-time retrieval results. We leverage a cross-encoder re-ranker to provide fine-grained pseudo labels over retrieval results and iteratively optimize query representations with the gradient descent method. Our theoretical analysis reveals that TouR can be viewed as a generalization of the classical Rocchio's algorithm for pseudo relevance feedback, and we present two variants leveraging psuedo labels as either hard binary or soft continuous labels. We first apply TouR on phrase retrieval with our proposed phrase re-ranker. On passage retrieval, we demonstrate its effectiveness with an off-the-shelf re-ranker. TouR improves the end-to-end open-domain QA accuracy significantly, as well as passage retrieval performance. Compared to re-ranker, TouR requires a smaller number of candidates, and achieves consistently better performance and runs up to 4x faster with our efficient implementation.

comment: 15 pages, 5 figures

♻ ☆ Reasoning over Different Types of Knowledge Graphs: Static, Temporal and Multi-Modal

Ke Liang, Lingyuan Meng, Meng Liu, Yue Liu, Wenxuan Tu, Siwei Wang, Sihang Zhou, Xinwang Liu, Fuchun Sun

Knowledge graph reasoning (KGR), aiming to deduce new facts from existing facts based on mined logic rules underlying knowledge graphs (KGs), has become a fast-growing research direction. It has been proven to significantly benefit the usage of KGs in many AI applications, such as question answering and recommendation systems, etc. According to the graph types, the existing KGR models can be roughly divided into three categories, i.e., static models, temporal models, and multi-modal models. The early works in this domain mainly focus on static KGR and tend to directly apply general knowledge graph embedding models to the reasoning task. However, these models are not suitable for more complex but practical tasks, such as inductive static KGR, temporal KGR, and multi-modal KGR. To this end, multiple works have been developed recently, but no survey papers and open-source repositories comprehensively summarize and discuss models in this important direction. To fill the gap, we conduct a survey for knowledge graph reasoning tracing from static to temporal and then to multi-modal KGs. Concretely, the preliminaries, summaries of KGR models, and typical datasets are introduced and discussed consequently. Moreover, we discuss the challenges and potential opportunities. The corresponding open-source repository is shared on GitHub: https://github.com/LIANGKE23/Awesome-Knowledge-Graph-Reasoning.

comment: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

☆ Personalized Decentralized Multi-Task Learning Over Dynamic Communication Graphs

Matin Mortaheb, Sennur Ulukus

Decentralized and federated learning algorithms face data heterogeneity as one of the biggest challenges, especially when users want to learn a specific task. Even when personalized headers are used concatenated to a shared network (PF-MTL), aggregating all the networks with a decentralized algorithm can result in performance degradation as a result of heterogeneity in the data. Our algorithm uses exchanged gradients to calculate the correlations among tasks automatically, and dynamically adjusts the communication graph to connect mutually beneficial tasks and isolate those that may negatively impact each other. This algorithm improves the learning performance and leads to faster convergence compared to the case where all clients are connected to each other regardless of their correlations. We conduct experiments on a synthetic Gaussian dataset and a large-scale celebrity attributes (CelebA) dataset. The experiment with the synthetic data illustrates that our proposed method is capable of detecting tasks that are positively and negatively correlated. Moreover, the results of the experiments with CelebA demonstrate that the proposed method may produce significantly faster training results than fully-connected networks.

☆ Contrastive Language-Vision AI Models Pretrained on Web-Scraped Multimodal Data Exhibit Sexual Objectification Bias

Robert Wolfe, Yiwei Yang, Bill Howe, Aylin Caliskan

Nine language-vision AI models trained on web scrapes with the Contrastive Language-Image Pretraining (CLIP) objective are evaluated for evidence of a bias studied by psychologists: the sexual objectification of girls and women, which occurs when a person's human characteristics are disregarded and the person is treated as a body or a collection of body parts. A first experiment uses standardized images of women from the Sexual OBjectification and EMotion Database, and finds that, commensurate with prior research in psychology, human characteristics are disassociated from images of objectified women: the model's recognition of emotional state is mediated by whether the subject is fully or partially clothed. Embedding association tests (EATs) return significant effect sizes for both anger (d >.8) and sadness (d >.5). A second experiment measures the effect in a representative application: an automatic image captioner (Antarctic Captions) includes words denoting emotion less than 50% as often for images of partially clothed women than for images of fully clothed women. A third experiment finds that images of female professionals (scientists, doctors, executives) are likely to be associated with sexual descriptions relative to images of male professionals. A fourth experiment shows that a prompt of "a [age] year old girl" generates sexualized images (as determined by an NSFW classifier) up to 73% of the time for VQGAN-CLIP (age 17), and up to 40% of the time for Stable Diffusion (ages 14 and 18); the corresponding rate for boys never surpasses 9%. The evidence indicates that language-vision AI models trained on automatically collected web scrapes learn biases of sexual objectification, which propagate to downstream applications.

comment: 15 pages, 2 figures

☆ Adapting to Latent Subgroup Shifts via Concepts and Proxies

Ibrahim Alabdulmohsin, Nicole Chiou, Alexander D'Amour, Arthur Gretton, Sanmi Koyejo, Matt J. Kusner, Stephen R. Pfohl, Olawale Salaudeen, Jessica Schrouff, Katherine Tsai

We address the problem of unsupervised domain adaptation when the source domain differs from the target domain because of a shift in the distribution of a latent subgroup. When this subgroup confounds all observed data, neither covariate shift nor label shift assumptions apply. We show that the optimal target predictor can be non-parametrically identified with the help of concept and proxy variables available only in the source domain, and unlabeled data from the target. The identification results are constructive, immediately suggesting an algorithm for estimating the optimal predictor in the target. For continuous observations, when this algorithm becomes impractical, we propose a latent variable model specific to the data generation process at hand. We show how the approach degrades as the size of the shift changes, and verify that it outperforms both covariate and label shift adjustment.

comment: Authors listed in alphabetical order

☆ Improving Narrative Relationship Embeddings by Training with Additional Inverse-Relationship Constraints

Mikolaj Figurski

We consider the problem of embedding character-entity relationships from the reduced semantic space of narratives, proposing and evaluating the assumption that these relationships hold under a reflection operation. We analyze this assumption and compare the approach to a baseline state-of-the-art model with a unique evaluation that simulates efficacy on a downstream clustering task with human-created labels. Although our model creates clusters that achieve Silhouette scores of -.084, outperforming the baseline -.227, our analysis reveals that the models approach the task much differently and perform well on very different examples. We conclude that our assumption might be useful for specific types of data and should be evaluated on a wider range of tasks.

☆ A Nearly Tight Bound for Fitting an Ellipsoid to Gaussian Random Points

Daniel M. Kane, Ilias Diakonikolas

We prove that for $c>0$ a sufficiently small universal constant that a random set of $c d^2/\log^4(d)$ independent Gaussian random points in $\mathbb{R}^d$ lie on a common ellipsoid with high probability. This nearly establishes a conjecture of~\cite{SaundersonCPW12}, within logarithmic factors. The latter conjecture has attracted significant attention over the past decade, due to its connections to machine learning and sum-of-squares lower bounds for certain statistical problems.

☆ A Theoretical Study of The Effects of Adversarial Attacks on Sparse Regression

Deepak Maurya, Jean Honorio

This paper analyzes $\ell_1$ regularized linear regression under the challenging scenario of having only adversarially corrupted data for training. We use the primal-dual witness paradigm to provide provable performance guarantees for the support of the estimated regression parameter vector to match the actual parameter. Our theoretical analysis shows the counter-intuitive result that an adversary can influence sample complexity by corrupting the irrelevant features, i.e., those corresponding to zero coefficients of the regression parameter vector, which, consequently, do not affect the dependent variable. As any adversarially robust algorithm has its limitations, our theoretical analysis identifies the regimes under which the learning algorithm and adversary can dominate over each other. It helps us to analyze these fundamental limits and address critical scientific questions of which parameters (like mutual incoherence, the maximum and minimum eigenvalue of the covariance matrix, and the budget of adversarial perturbation) play a role in the high or low probability of success of the LASSO algorithm. Also, the derived sample complexity is logarithmic with respect to the size of the regression parameter vector, and our theoretical claims are validated by empirical analysis on synthetic and real-world datasets.

comment: first version

☆ A Seven-Layer Model for Standardising AI Fairness Assessment

Avinash Agarwal, Harsh Agarwal

Problem statement: Standardisation of AI fairness rules and benchmarks is challenging because AI fairness and other ethical requirements depend on multiple factors such as context, use case, type of the AI system, and so on. In this paper, we elaborate that the AI system is prone to biases at every stage of its lifecycle, from inception to its usage, and that all stages require due attention for mitigating AI bias. We need a standardised approach to handle AI fairness at every stage. Gap analysis: While AI fairness is a hot research topic, a holistic strategy for AI fairness is generally missing. Most researchers focus only on a few facets of AI model-building. Peer review shows excessive focus on biases in the datasets, fairness metrics, and algorithmic bias. In the process, other aspects affecting AI fairness get ignored. The solution proposed: We propose a comprehensive approach in the form of a novel seven-layer model, inspired by the Open System Interconnection (OSI) model, to standardise AI fairness handling. Despite the differences in the various aspects, most AI systems have similar model-building stages. The proposed model splits the AI system lifecycle into seven abstraction layers, each corresponding to a well-defined AI model-building or usage stage. We also provide checklists for each layer and deliberate on potential sources of bias in each layer and their mitigation methodologies. This work will facilitate layer-wise standardisation of AI fairness rules and benchmarking parameters.

comment: 13 pages

☆ Vulnerabilities of Deep Learning-Driven Semantic Communications to Backdoor (Trojan) Attacks

Yalin E. Sagduyu, Tugba Erpek, Sennur Ulukus, Aylin Yener

This paper highlights vulnerabilities of deep learning-driven semantic communications to backdoor (Trojan) attacks. Semantic communications aims to convey a desired meaning while transferring information from a transmitter to its receiver. An encoder-decoder pair that is represented by two deep neural networks (DNNs) as part of an autoencoder is trained to reconstruct signals such as images at the receiver by transmitting latent features of small size over a limited number of channel uses. In the meantime, another DNN of a semantic task classifier at the receiver is jointly trained with the autoencoder to check the meaning conveyed to the receiver. The complex decision space of the DNNs makes semantic communications susceptible to adversarial manipulations. In a backdoor (Trojan) attack, the adversary adds triggers to a small portion of training samples and changes the label to a target label. When the transfer of images is considered, the triggers can be added to the images or equivalently to the corresponding transmitted or received signals. In test time, the adversary activates these triggers by providing poisoned samples as input to the encoder (or decoder) of semantic communications. The backdoor attack can effectively change the semantic information transferred for the poisoned input samples to a target meaning. As the performance of semantic communications improves with the signal-to-noise ratio and the number of channel uses, the success of the backdoor attack increases as well. Also, increasing the Trojan ratio in training data makes the attack more successful. In the meantime, the effect of this attack on the unpoisoned input samples remains limited. Overall, this paper shows that the backdoor attack poses a serious threat to semantic communications and presents novel design guidelines to preserve the meaning of transferred information in the presence of backdoor attacks.

☆ Deep Reinforcement Learning for Trajectory Path Planning and Distributed Inference in Resource-Constrained UAV Swarms

Marwan Dhuheir, Emna Baccour, Aiman Erbad, Sinan Sabeeh Al-Obaidi, Mounir Hamdi

The deployment flexibility and maneuverability of Unmanned Aerial Vehicles (UAVs) increased their adoption in various applications, such as wildfire tracking, border monitoring, etc. In many critical applications, UAVs capture images and other sensory data and then send the captured data to remote servers for inference and data processing tasks. However, this approach is not always practical in real-time applications due to the connection instability, limited bandwidth, and end-to-end latency. One promising solution is to divide the inference requests into multiple parts (layers or segments), with each part being executed in a different UAV based on the available resources. Furthermore, some applications require the UAVs to traverse certain areas and capture incidents; thus, planning their paths becomes critical particularly, to reduce the latency of making the collaborative inference process. Specifically, planning the UAVs trajectory can reduce the data transmission latency by communicating with devices in the same proximity while mitigating the transmission interference. This work aims to design a model for distributed collaborative inference requests and path planning in a UAV swarm while respecting the resource constraints due to the computational load and memory usage of the inference requests. The model is formulated as an optimization problem and aims to minimize latency. The formulated problem is NP-hard so finding the optimal solution is quite complex; thus, this paper introduces a real-time and dynamic solution for online applications using deep reinforcement learning. We conduct extensive simulations and compare our results to the-state-of-the-art studies demonstrating that our model outperforms the competing models.

comment: accepted journal paper at IEEE Internet of Things Journal

☆ Free-Rider Games for Federated Learning with Selfish Clients in NextG Wireless Networks

Yalin E. Sagduyu

This paper presents a game theoretic framework for participation and free-riding in federated learning (FL), and determines the Nash equilibrium strategies when FL is executed over wireless links. To support spectrum sensing for NextG communications, FL is used by clients, namely spectrum sensors with limited training datasets and computation resources, to train a wireless signal classifier while preserving privacy. In FL, a client may be free-riding, i.e., it does not participate in FL model updates, if the computation and transmission cost for FL participation is high, and receives the global model (learned by other clients) without incurring a cost. However, the free-riding behavior may potentially decrease the global accuracy due to lack of contribution to global model learning. This tradeoff leads to a non-cooperative game where each client aims to individually maximize its utility as the difference between the global model accuracy and the cost of FL participation. The Nash equilibrium strategies are derived for free-riding probabilities such that no client can unilaterally increase its utility given the strategies of its opponents remain the same. The free-riding probability increases with the FL participation cost and the number of clients, and a significant optimality gap exists in Nash equilibrium with respect to the joint optimization for all clients. The optimality gap increases with the number of clients and the maximum gap is evaluated as a function of the cost. These results quantify the impact of free-riding on the resilience of FL in NextG networks and indicate operational modes for FL participation.

☆ Similarity Contrastive Estimation for Image and Video Soft Contrastive Self-Supervised Learning WACV 2023

Julien Denize, Jaonary Rabarisoa, Astrid Orcesi, Romain Hérault

Contrastive representation learning has proven to be an effective self-supervised learning method for images and videos. Most successful approaches are based on Noise Contrastive Estimation (NCE) and use different views of an instance as positives that should be contrasted with other instances, called negatives, that are considered as noise. However, several instances in a dataset are drawn from the same distribution and share underlying semantic information. A good data representation should contain relations between the instances, or semantic similarity and dissimilarity, that contrastive learning harms by considering all negatives as noise. To circumvent this issue, we propose a novel formulation of contrastive learning using semantic similarity between instances called Similarity Contrastive Estimation (SCE). Our training objective is a soft contrastive one that brings the positives closer and estimates a continuous distribution to push or pull negative instances based on their learned similarities. We validate empirically our approach on both image and video representation learning. We show that SCE performs competitively with the state of the art on the ImageNet linear evaluation protocol for fewer pretraining epochs and that it generalizes to several downstream image tasks. We also show that SCE reaches state-of-the-art results for pretraining video representation and that the learned representation can generalize to video downstream tasks.

comment: Extended version of our WACV 2023 paper to video self-supervised learning

☆ Robust Path Selection in Software-defined WANs using Deep Reinforcement Learning

Shahrooz Pouryousef, Lixin Gao, Don Towsley

In the context of an efficient network traffic engineering process where the network continuously measures a new traffic matrix and updates the set of paths in the network, an automated process is required to quickly and efficiently identify when and what set of paths should be used. Unfortunately, the burden of finding the optimal solution for the network updating process in each given time interval is high since the computation complexity of optimization approaches using linear programming increases significantly as the size of the network increases. In this paper, we use deep reinforcement learning to derive a data-driven algorithm that does the path selection in the network considering the overhead of route computation and path updates. Our proposed scheme leverages information about past network behavior to identify a set of robust paths to be used for multiple future time intervals to avoid the overhead of updating the forwarding behavior of routers frequently. We compare the results of our approach to other traffic engineering solutions through extensive simulations across real network topologies. Our results demonstrate that our scheme fares well by a factor of 40% with respect to reducing link utilization compared to traditional TE schemes such as ECMP. Our scheme provides a slightly higher link utilization (around 25%) compared to schemes that only minimize link utilization and do not care about path updating overhead.

☆ Efficient First-order Methods for Convex Optimization with Strongly Convex Function Constraints

Zhenwei Lin, Qi Deng

Convex function constrained optimization has received growing research interests lately. For a special convex problem which has strongly convex function constraints, we develop a new accelerated primal-dual first-order method that obtains an $\Ocal(1/\sqrt{\vep})$ complexity bound, improving the $\Ocal(1/{\vep})$ result for the state-of-the-art first-order methods. The key ingredient to our development is some novel techniques to progressively estimate the strong convexity of the Lagrangian function, which enables adaptive step-size selection and faster convergence performance. In addition, we show that the complexity is further improvable in terms of the dependence on some problem parameter, via a restart scheme that calls the accelerated method repeatedly. As an application, we consider sparsity-inducing constrained optimization which has a separable convex objective and a strongly convex loss constraint. In addition to achieving fast convergence, we show that the restarted method can effectively identify the sparsity pattern (active-set) of the optimal solution in finite steps. To the best of our knowledge, this is the first active-set identification result for sparsity-inducing constrained optimization.

comment: 27 pages, 3 figures

☆ Generating music with sentiment using Transformer-GANs

Pedro Neves, Jose Fornari, João Florindo

The field of Automatic Music Generation has seen significant progress thanks to the advent of Deep Learning. However, most of these results have been produced by unconditional models, which lack the ability to interact with their users, not allowing them to guide the generative process in meaningful and practical ways. Moreover, synthesizing music that remains coherent across longer timescales while still capturing the local aspects that make it sound ``realistic'' or ``human-like'' is still challenging. This is due to the large computational requirements needed to work with long sequences of data, and also to limitations imposed by the training schemes that are often employed. In this paper, we propose a generative model of symbolic music conditioned by data retrieved from human sentiment. The model is a Transformer-GAN trained with labels that correspond to different configurations of the valence and arousal dimensions that quantitatively represent human affective states. We try to tackle both of the problems above by employing an efficient linear version of Attention and using a Discriminator both as a tool to improve the overall quality of the generated music and its ability to follow the conditioning signals.

☆ Towards dynamic stability analysis of sustainable power grids using graph neural networks

Christian Nauck, Michael Lindner, Konstantin Schürholt, Frank Hellmann

To mitigate climate change, the share of renewable needs to be increased. Renewable energies introduce new challenges to power grids due to decentralization, reduced inertia and volatility in production. The operation of sustainable power grids with a high penetration of renewable energies requires new methods to analyze the dynamic stability. We provide new datasets of dynamic stability of synthetic power grids and find that graph neural networks (GNNs) are surprisingly effective at predicting the highly non-linear target from topological information only. To illustrate the potential to scale to real-sized power grids, we demonstrate the successful prediction on a Texan power grid model.

comment: main section: 4 pages, 3 figures

☆ Lifelong Reinforcement Learning with Modulating Masks

Eseoghene Ben-Iwhiwhu, Saptarshi Nath, Praveen K. Pilly, Soheil Kolouri, Andrea Soltoggio

Lifelong learning aims to create AI systems that continuously and incrementally learn during a lifetime, similar to biological learning. Attempts so far have met problems, including catastrophic forgetting, interference among tasks, and the inability to exploit previous knowledge. While considerable research has focused on learning multiple input distributions, typically in classification, lifelong reinforcement learning (LRL) must also deal with variations in the state and transition distributions, and in the reward functions. Modulating masks, recently developed for classification, are particularly suitable to deal with such a large spectrum of task variations. In this paper, we adapted modulating masks to work with deep LRL, specifically PPO and IMPALA agents. The comparison with LRL baselines in both discrete and continuous RL tasks shows competitive performance. We further investigated the use of a linear combination of previously learned masks to exploit previous knowledge when learning new tasks: not only is learning faster, the algorithm solves tasks that we could not otherwise solve from scratch due to extremely sparse rewards. The results suggest that RL with modulating masks is a promising approach to lifelong learning, to the composition of knowledge to learn increasingly complex tasks, and to knowledge reuse for efficient and faster learning.

☆ On Reinforcement Learning for the Game of 2048

Hung Guei

2048 is a single-player stochastic puzzle game. This intriguing and addictive game has been popular worldwide and has attracted researchers to develop game-playing programs. Due to its simplicity and complexity, 2048 has become an interesting and challenging platform for evaluating the effectiveness of machine learning methods. This dissertation conducts comprehensive research on reinforcement learning and computer game algorithms for 2048. First, this dissertation proposes optimistic temporal difference learning, which significantly improves the quality of learning by employing optimistic initialization to encourage exploration for 2048. Furthermore, based on this approach, a state-of-the-art program for 2048 is developed, which achieves the highest performance among all learning-based programs, namely an average score of 625377 points and a rate of 72% for reaching 32768-tiles. Second, this dissertation investigates several techniques related to 2048, including the n-tuple network ensemble learning, Monte Carlo tree search, and deep reinforcement learning. These techniques are promising for further improving the performance of the current state-of-the-art program. Finally, this dissertation discusses pedagogical applications related to 2048 by proposing course designs and summarizing the teaching experience. The proposed course designs use 2048-like games as materials for beginners to learn reinforcement learning and computer game algorithms. The courses have been successfully applied to graduate-level students and received well by student feedback.

comment: A Ph.D. dissertation submitted to Institute of Computer Science and Engineering, National Yang Ming Chiao Tung University

☆ Is it worth it? An experimental comparison of six deep- and classical machine learning methods for unsupervised anomaly detection in time series

Ferdinand Rewicki, Joachim Denzler, Julia Niebling

The detection of anomalies in time series data is crucial in a wide range of applications, such as system monitoring, health care or cyber security. While the vast number of available methods makes selecting the right method for a certain application hard enough, different methods have different strengths, e.g. regarding the type of anomalies they are able to find. In this work, we compare six unsupervised anomaly detection methods with different complexities to answer the questions: Are the more complex methods usually performing better? And are there specific anomaly types that those method are tailored to? The comparison is done on the UCR anomaly archive, a recent benchmark dataset for anomaly detection. We compare the six methods by analyzing the experimental results on a dataset- and anomaly type level after tuning the necessary hyperparameter for each method. Additionally we examine the ability of individual methods to incorporate prior knowledge about the anomalies and analyse the differences of point-wise and sequence wise features. We show with broad experiments, that the classical machine learning methods show a superior performance compared to the deep learning methods across a wide range of anomaly types.

comment: 15 Pages, The repository to reproduce the results is available at https://gitlab.com/dlr-dw/is-it-worth-it-benchmark

☆ Revisiting Residual Networks for Adversarial Robustness: An Architectural Perspective

Shihua Huang, Zhichao Lu, Kalyanmoy Deb, Vishnu Naresh Boddeti

Efforts to improve the adversarial robustness of convolutional neural networks have primarily focused on developing more effective adversarial training methods. In contrast, little attention was devoted to analyzing the role of architectural elements (such as topology, depth, and width) on adversarial robustness. This paper seeks to bridge this gap and present a holistic study on the impact of architectural design on adversarial robustness. We focus on residual networks and consider architecture design at the block level, i.e., topology, kernel size, activation, and normalization, as well as at the network scaling level, i.e., depth and width of each block in the network. In both cases, we first derive insights through systematic ablative experiments. Then we design a robust residual block, dubbed RobustResBlock, and a compound scaling rule, dubbed RobustScaling, to distribute depth and width at the desired FLOP count. Finally, we combine RobustResBlock and RobustScaling and present a portfolio of adversarially robust residual networks, RobustResNets, spanning a broad spectrum of model capacities. Experimental validation across multiple datasets and adversarial attacks demonstrate that RobustResNets consistently outperform both the standard WRNs and other existing robust architectures, achieving state-of-the-art AutoAttack robust accuracy of 61.1% without additional data and 63.7% with 500K external data while being $2\times$ more compact in terms of parameters. Code is available at \url{ https://github.com/zhichao-lu/robust-residual-network}

☆ LogAnMeta: Log Anomaly Detection Using Meta Learning

Abhishek Sarkar, Tanmay Sen, Srimanta Kundu, Arijit Sarkar, Abdul Wazed

Modern telecom systems are monitored with performance and system logs from multiple application layers and components. Detecting anomalous events from these logs is key to identify security breaches, resource over-utilization, critical/fatal errors, etc. Current supervised log anomaly detection frameworks tend to perform poorly on new types or signatures of anomalies with few or unseen samples in the training data. In this work, we propose a meta-learning-based log anomaly detection framework (LogAnMeta) for detecting anomalies from sequence of log events with few samples. LoganMeta train a hybrid few-shot classifier in an episodic manner. The experimental results demonstrate the efficacy of our proposed method

☆ SoK: Let The Privacy Games Begin! A Unified Treatment of Data Inference Privacy in Machine Learning

Ahmed Salem, Giovanni Cherubin, David Evans, Boris Köpf, Andrew Paverd, Anshuman Suri, Shruti Tople, Santiago Zanella-Béguelin

Deploying machine learning models in production may allow adversaries to infer sensitive information about training data. There is a vast literature analyzing different types of inference risks, ranging from membership inference to reconstruction attacks. Inspired by the success of games (i.e., probabilistic experiments) to study security properties in cryptography, some authors describe privacy inference risks in machine learning using a similar game-based style. However, adversary capabilities and goals are often stated in subtly different ways from one presentation to the other, which makes it hard to relate and compose results. In this paper, we present a game-based framework to systematize the body of knowledge on privacy inference risks in machine learning.

☆ The Ties that matter: From the perspective of Similarity Measure in Online Social Networks

Soumita Das, Anupam Biswas

Online Social Networks have embarked on the importance of connection strength measures which has a broad array of applications such as, analyzing diffusion behaviors, community detection, link predictions, recommender systems. Though there are some existing connection strength measures, the density that a connection shares with it's neighbors and the directionality aspect has not received much attention. In this paper, we have proposed an asymmetric edge similarity measure namely, Neighborhood Density-based Edge Similarity (NDES) which provides a fundamental support to derive the strength of connection. The time complexity of NDES is $O(nk^2)$. An application of NDES for community detection in social network is shown. We have considered a similarity based community detection technique and substituted its similarity measure with NDES. The performance of NDES is evaluated on several small real-world datasets in terms of the effectiveness in detecting communities and compared with three widely used similarity measures. Empirical results show NDES enables detecting comparatively better communities both in terms of accuracy and quality.

comment: To be published in MINDS-2021

☆ Joint Embedding of 2D and 3D Networks for Medical Image Anomaly Detection

Inha Kang, Jinah Park

Obtaining ground truth data in medical imaging has difficulties due to the fact that it requires a lot of annotating time from the experts in the field. Also, when trained with supervised learning, it detects only the cases included in the labels. In real practice, we want to also open to other possibilities than the named cases while examining the medical images. As a solution, the need for anomaly detection that can detect and localize abnormalities by learning the normal characteristics using only normal images is emerging. With medical image data, we can design either 2D or 3D networks of self-supervised learning for anomaly detection task. Although 3D networks, which learns 3D structures of the human body, show good performance in 3D medical image anomaly detection, they cannot be stacked in deeper layers due to memory problems. While 2D networks have advantage in feature detection, they lack 3D context information. In this paper, we develop a method for combining the strength of the 3D network and the strength of the 2D network through joint embedding. We also propose the pretask of self-supervised learning to make it possible for the networks to learn efficiently. Through the experiments, we show that the proposed method achieves better performance in both classification and segmentation tasks compared to the SoTA method.

comment: 15 pages, 11 figures

☆ DCC: A Cascade based Approach to Detect Communities in Social Networks

Soumita Das, Anupam Biswas, Akrati Saxena

Community detection in Social Networks is associated with finding and grouping the most similar nodes inherent in the network. These similar nodes are identified by computing tie strength. Stronger ties indicates higher proximity shared by connected node pairs. This work is motivated by Granovetter's argument that suggests that strong ties lies within densely connected nodes and the theory that community cores in real-world networks are densely connected. In this paper, we have introduced a novel method called \emph{Disjoint Community detection using Cascades (DCC)} which demonstrates the effectiveness of a new local density based tie strength measure on detecting communities. Here, tie strength is utilized to decide the paths followed for propagating information. The idea is to crawl through the tuple information of cascades towards the community core guided by increasing tie strength. Considering the cascade generation step, a novel preferential membership method has been developed to assign community labels to unassigned nodes. The efficacy of $DCC$ has been analyzed based on quality and accuracy on several real-world datasets and baseline community detection algorithms.

comment: To be published in CHSN-2022

☆ A Memetic Algorithm with Reinforcement Learning for Sociotechnical Production Scheduling

Felix Grumbach, Nour Eldin Alaa Badr, Pascal Reusch, Sebastian Trojahn

The following article presents a memetic algorithm with applying deep reinforcement learning (DRL) for solving practically oriented dual resource constrained flexible job shop scheduling problems (DRC-FJSSP). In recent years, there has been extensive research on DRL techniques, but without considering realistic, flexible and human-centered shopfloors. A research gap can be identified in the context of make-to-order oriented discontinuous manufacturing as it is often represented in medium-size companies with high service levels. From practical industry projects in this domain, we recognize requirements to depict flexible machines, human workers and capabilities, setup and processing operations, material arrival times, complex job paths with parallel tasks for bill of material (BOM) manufacturing, sequence-depended setup times and (partially) automated tasks. On the other hand, intensive research has been done on metaheuristics in the context of DRC-FJSSP. However, there is a lack of suitable and generic scheduling methods that can be holistically applied in sociotechnical production and assembly processes. In this paper, we first formulate an extended DRC-FJSSP induced by the practical requirements mentioned. Then we present our proposed hybrid framework with parallel computing for multicriteria optimization. Through numerical experiments with real-world data, we confirm that the framework generates feasible schedules efficiently and reliably. Utilizing DRL instead of random operations leads to better results and outperforms traditional approaches.

☆ Minimizing Worst-Case Violations of Neural Networks

Rahul Nellikkath, Spyros Chatzivasileiadis

Machine learning (ML) algorithms are remarkably good at approximating complex non-linear relationships. Most ML training processes, however, are designed to deliver ML tools with good average performance, but do not offer any guarantees about their worst-case estimation error. For safety-critical systems such as power systems, this places a major barrier for their adoption. So far, approaches could determine the worst-case violations of only trained ML algorithms. To the best of our knowledge, this is the first paper to introduce a neural network training procedure designed to achieve both a good average performance and minimum worst-case violations. Using the Optimal Power Flow (OPF) problem as a guiding application, our approach (i) introduces a framework that reduces the worst-case generation constraint violations during training, incorporating them as a differentiable optimization layer; and (ii) presents a neural network sequential learning architecture to significantly accelerate it. We demonstrate the proposed architecture on four different test systems ranging from 39 buses to 162 buses, for both AC-OPF and DC-OPF applications.

☆ SPT: Semi-Parametric Prompt Tuning for Multitask Prompted Learning

M Saiful Bari, Aston Zhang, Shuai Zheng, Xingjian Shi, Yi Zhu, Shafiq Joty, Mu Li

Pre-trained large language models can efficiently interpolate human-written prompts in a natural way. Multitask prompted learning can help generalization through a diverse set of tasks at once, thus enhancing the potential for more effective downstream fine-tuning. To perform efficient multitask-inference in the same batch, parameter-efficient fine-tuning methods such as prompt tuning have been proposed. However, the existing prompt tuning methods may lack generalization. We propose SPT, a semi-parametric prompt tuning method for multitask prompted learning. The novel component of SPT is a memory bank from where memory prompts are retrieved based on discrete prompts. Extensive experiments, such as (i) fine-tuning a full language model with SPT on 31 different tasks from 8 different domains and evaluating zero-shot generalization on 9 heldout datasets under 5 NLP task categories and (ii) pretraining SPT on the GLUE datasets and evaluating fine-tuning on the SuperGLUE datasets, demonstrate effectiveness of SPT.

☆ Ensemble learning techniques for intrusion detection system in the context of cybersecurity

Andricson Abeline Moreira, Carlos A. C. Tojeiro, Carlos J. Reis, Gustavo Henrique Massaro, Igor Andrade Brito e Kelton A. P. da Costa

Recently, there has been an interest in improving the resources available in Intrusion Detection System (IDS) techniques. In this sense, several studies related to cybersecurity show that the environment invasions and information kidnapping are increasingly recurrent and complex. The criticality of the business involving operations in an environment using computing resources does not allow the vulnerability of the information. Cybersecurity has taken on a dimension within the universe of indispensable technology in corporations, and the prevention of risks of invasions into the environment is dealt with daily by Security teams. Thus, the main objective of the study was to investigate the Ensemble Learning technique using the Stacking method, supported by the Support Vector Machine (SVM) and k-Nearest Neighbour (kNN) algorithms aiming at an optimization of the results for DDoS attack detection. For this, the Intrusion Detection System concept was used with the application of the Data Mining and Machine Learning Orange tool to obtain better results

comment: in Portuguese language. CIACA - Conferencia Ibero-Americana Computa\c{c}\~ao Aplicada 2022 Proceedings

☆ A Survey of Mix-based Data Augmentation: Taxonomy, Methods, Applications, and Explainability

Chengtai Cao, Fan Zhou, Yurou Dai, Jianping Wang

Data augmentation (DA) is indispensable in modern machine learning and deep neural networks. The basic idea of DA is to construct new training data to improve the model's generalization by adding slightly disturbed versions of existing data or synthesizing new data. In this work, we review a small but essential subset of DA -- Mix-based Data Augmentation (MixDA) that generates novel samples by mixing multiple examples. Unlike conventional DA approaches based on a single-sample operation or requiring domain knowledge, MixDA is more general in creating a broad spectrum of new data and has received increasing attention in the community. We begin with proposing a new taxonomy classifying MixDA into, Mixup-based, Cutmix-based, and hybrid approaches according to a hierarchical view of the data mix. Various MixDA techniques are then comprehensively reviewed in a more fine-grained way. Owing to its generalization, MixDA has penetrated a variety of applications which are also completely reviewed in this work. We also examine why MixDA works from different aspects of improving model performance, generalization, and calibration while explaining the model behavior based on the properties of MixDA. Finally, we recapitulate the critical findings and fundamental challenges of current MixDA studies, and outline the potential directions for future works. Different from previous related works that summarize the DA approaches in a specific domain (e.g., images or natural language processing) or only review a part of MixDA studies, we are the first to provide a systematical survey of MixDA in terms of its taxonomy, methodology, applications, and explainability. This work can serve as a roadmap to MixDA techniques and application reviews while providing promising directions for researchers interested in this exciting area.

☆ Hyperparameters in Contextual RL are Highly Situational

Theresa Eimer, Carolin Benjamins, Marius Lindauer

Although Reinforcement Learning (RL) has shown impressive results in games and simulation, real-world application of RL suffers from its instability under changing environment conditions and hyperparameters. We give a first impression of the extent of this instability by showing that the hyperparameters found by automatic hyperparameter optimization (HPO) methods are not only dependent on the problem at hand, but even on how well the state describes the environment dynamics. Specifically, we show that agents in contextual RL require different hyperparameters if they are shown how environmental factors change. In addition, finding adequate hyperparameter configurations is not equally easy for both settings, further highlighting the need for research into how hyperparameters influence learning and generalization in RL.

☆ Prompt-Augmented Linear Probing: Scaling Beyond The Limit of Few-shot In-Context Learners AAAI 2023

Hyunsoo Cho, Hyuhng Joon Kim, Junyeob Kim, Sang-Woo Lee, Sang-goo Lee, Kang Min Yoo, Taeuk Kim

Through in-context learning (ICL), large-scale language models are effective few-shot learners without additional model fine-tuning. However, the ICL performance does not scale well with the number of available training samples as it is limited by the inherent input length constraint of the underlying language model. Meanwhile, many studies have revealed that language models are also powerful feature extractors, allowing them to be utilized in a black-box manner and enabling the linear probing paradigm, where lightweight discriminators are trained on top of the pre-extracted input representations. This paper proposes prompt-augmented linear probing (PALP), a hybrid of linear probing and ICL, which leverages the best of both worlds. PALP inherits the scalability of linear probing and the capability of enforcing language models to derive more meaningful representations via tailoring input into a more conceivable form. Throughout in-depth investigations on various datasets, we verified that PALP significantly enhances the input representations closing the gap between ICL in the data-hungry scenario and fine-tuning in the data-abundant scenario with little training overhead, potentially making PALP a strong alternative in a black-box scenario.

comment: AAAI 2023

☆ 5G Long-Term and Large-Scale Mobile Traffic Forecasting

Ufuk Uyan, M. Tugberk Isyapar, Mahiye Uluyagmur Ozturk

It is crucial for the service provider to comprehend and forecast mobile traffic in large-scale cellular networks in order to govern and manage mechanisms for base station placement, load balancing, and network planning. The purpose of this article is to extract and simulate traffic patterns from more than 14,000 cells that have been installed in different metropolitan areas. To do this, we create, implement, and assess a method in which cells are first categorized by their point of interest and then clustered based on the temporal distribution of cells in each region. The proposed model has been tested using real-world 5G mobile traffic datasets collected over 31 weeks in various cities. We found that our proposed model performed well in predicting mobile traffic patterns up to 2 weeks in advance. Our model outperformed the base model in most areas of interest and generally achieved up to 15\% less prediction error compared to the na\"ive approach. This indicates that our approach is effective in predicting mobile traffic patterns in large-scale cellular networks.

☆ Temporal Disaggregation of the Cumulative Grass Growth

Thomas Guyet, Laurent Spillemaecker, Simon Malinowski, Anne-Isabelle Graux

Information on the grass growth over a year is essential for some models simulating the use of this resource to feed animals on pasture or at barn with hay or grass silage. Unfortunately, this information is rarely available. The challenge is to reconstruct grass growth from two sources of information: usual daily climate data (rainfall, radiation, etc.) and cumulative growth over the year. We have to be able to capture the effect of seasonal climatic events which are known to distort the growth curve within the year. In this paper, we formulate this challenge as a problem of disaggregating the cumulative growth into a time series. To address this problem, our method applies time series forecasting using climate information and grass growth from previous time steps. Several alternatives of the method are proposed and compared experimentally using a database generated from a grassland process-based model. The results show that our method can accurately reconstruct the time series, independently of the use of the cumulative growth information.

☆ PABAU: Privacy Analysis of Biometric API Usage

Feiyang Tang

Biometric data privacy is becoming a major concern for many organizations in the age of big data, particularly in the ICT sector, because it may be easily exploited in apps. Most apps utilize biometrics by accessing common application programming interfaces (APIs); hence, we aim to categorize their usage. The categorization based on behavior may be closely correlated with the sensitive processing of a user's biometric data, hence highlighting crucial biometric data privacy assessment concerns. We propose PABAU, Privacy Analysis of Biometric API Usage. PABAU learns semantic features of methods in biometric APIs and uses them to detect and categorize the usage of biometric API implementation in the software according to their privacy-related behaviors. This technique bridges the communication and background knowledge gap between technical and non-technical individuals in organizations by providing an automated method for both parties to acquire a rapid understanding of the essential behaviors of biometric API in apps, as well as future support to data protection officers (DPO) with legal documentation, such as conducting a Data Protection Impact Assessment (DPIA).

comment: Accepted by The 8th IEEE International Conference on Privacy Computing (PriComp 2022)

☆ VCNet: A self-explaining model for realistic counterfactual generation

Victor Guyomard, Françoise Fessant, Thomas Guyet, Tassadit Bouadi, Alexandre Termier

Counterfactual explanation is a common class of methods to make local explanations of machine learning decisions. For a given instance, these methods aim to find the smallest modification of feature values that changes the predicted decision made by a machine learning model. One of the challenges of counterfactual explanation is the efficient generation of realistic counterfactuals. To address this challenge, we propose VCNet-Variational Counter Net-a model architecture that combines a predictor and a counterfactual generator that are jointly trained, for regression or classification tasks. VCNet is able to both generate predictions, and to generate counterfactual explanations without having to solve another minimisation problem. Our contribution is the generation of counterfactuals that are close to the distribution of the predicted class. This is done by learning a variational autoencoder conditionally to the output of the predictor in a join-training fashion. We present an empirical evaluation on tabular datasets and across several interpretability metrics. The results are competitive with the state-of-the-art method.

☆ Greenhouse gases emissions: estimating corporate non-reported emissions using interpretable machine learning

Jeremi Assael, Thibaut Heurtebize, Laurent Carlier, François Soupé

As of 2022, greenhouse gases (GHG) emissions reporting and auditing are not yet compulsory for all companies and methodologies of measurement and estimation are not unified. We propose a machine learning-based model to estimate scope 1 and scope 2 GHG emissions of companies not reporting them yet. Our model, specifically designed to be transparent and completely adapted to this use case, is able to estimate emissions for a large universe of companies. It shows good out-of-sample global performances as well as good out-of-sample granular performances when evaluating it by sectors, by countries or by revenues buckets. We also compare our results to those of other providers and find our estimates to be more accurate. Thanks to the proposed explainability tools using Shapley values, our model is fully interpretable, the user being able to understand which factors split explain the GHG emissions for each particular company.

☆ Predicting the Score of Atomic Candidate OWL Class Axioms

Ali Ballout, Andrea G B Tettamanzi, Célia da Costa Pereira

Candidate axiom scoring is the task of assessing the acceptability of a candidate axiom against the evidence provided by known facts or data. The ability to score candidate axioms reliably is required for automated schema or ontology induction, but it can also be valuable for ontology and/or knowledge graph validation. Accurate axiom scoring heuristics are often computationally expensive, which is an issue if you wish to use them in iterative search techniques like level-wise generate-and-test or evolutionary algorithms, which require scoring a large number of candidate axioms. We address the problem of developing a predictive model as a substitute for reasoning that predicts the possibility score of candidate class axioms and is quick enough to be employed in such situations. We use a semantic similarity measure taken from an ontology's subsumption structure for this purpose. We show that the approach provided in this work can accurately learn the possibility scores of candidate OWL class axioms and that it can do so for a variety of OWL class axioms.

☆ Crab: Learning Certifiably Fair Predictive Models in the Presence of Selection Bias

Jiongli Zhu, Nazanin Sabri, Sainyam Galhotra, Babak Salimi

A recent explosion of research focuses on developing methods and tools for building fair predictive models. However, most of this work relies on the assumption that the training and testing data are representative of the target population on which the model will be deployed. However, real-world training data often suffer from selection bias and are not representative of the target population for many reasons, including the cost and feasibility of collecting and labeling data, historical discrimination, and individual biases. In this paper, we introduce a new framework for certifying and ensuring the fairness of predictive models trained on biased data. We take inspiration from query answering over incomplete and inconsistent databases to present and formalize the problem of consistent range approximation (CRA) of answers to queries about aggregate information for the target population. We aim to leverage background knowledge about the data collection process, biased data, and limited or no auxiliary data sources to compute a range of answers for aggregate queries over the target population that are consistent with available information. We then develop methods that use CRA of such aggregate queries to build predictive models that are certifiably fair on the target population even when no external information about that population is available during training. We evaluate our methods on real data and demonstrate improvements over state of the art. Significantly, we show that enforcing fairness using our methods can lead to predictive models that are not only fair, but more accurate on the target population.

☆ Towards Rapid Prototyping and Comparability in Active Learning for Deep Object Detection

Tobias Riedlinger, Marius Schubert, Karsten Kahl, Hanno Gottschalk, Matthias Rottmann

Active learning as a paradigm in deep learning is especially important in applications involving intricate perception tasks such as object detection where labels are difficult and expensive to acquire. Development of active learning methods in such fields is highly computationally expensive and time consuming which obstructs the progression of research and leads to a lack of comparability between methods. In this work, we propose and investigate a sandbox setup for rapid development and transparent evaluation of active learning in deep object detection. Our experiments with commonly used configurations of datasets and detection architectures found in the literature show that results obtained in our sandbox environment are representative of results on standard configurations. The total compute time to obtain results and assess the learning behavior can thereby be reduced by factors of up to 14 when comparing with Pascal VOC and up to 32 when comparing with BDD100k. This allows for testing and evaluating data acquisition and labeling strategies in under half a day and contributes to the transparency and development speed in the field of active learning for object detection.

comment: 17 pages, 12 figures, 9 tables

☆ Complete the Missing Half: Augmenting Aggregation Filtering with Diversification for Graph Convolutional Neural Networks NeurIPS 2022

Sitao Luan, Mingde Zhao, Chenqing Hua, Xiao-Wen Chang, Doina Precup

The core operation of current Graph Neural Networks (GNNs) is the aggregation enabled by the graph Laplacian or message passing, which filters the neighborhood information of nodes. Though effective for various tasks, in this paper, we show that they are potentially a problematic factor underlying all GNN models for learning on certain datasets, as they force the node representations similar, making the nodes gradually lose their identity and become indistinguishable. Hence, we augment the aggregation operations with their dual, i.e. diversification operators that make the node more distinct and preserve the identity. Such augmentation replaces the aggregation with a two-channel filtering process that, in theory, is beneficial for enriching the node representations. In practice, the proposed two-channel filters can be easily patched on existing GNN methods with diverse training strategies, including spectral and spatial (message passing) methods. In the experiments, we observe desired characteristics of the models and significant performance boost upon the baselines on 9 node classification tasks.

comment: Accepted as Oral Presentation at NeurIPS 2022 New Frontiers in Graph Learning Workshop (NeurIPS GLFrontiers 2022)

☆ Semi-Supervised Bifold Teacher-Student Learning for Indoor Presence Detection Under Time-Varying CSI

Li-Hsiang Shen, Kai-Jui Chen, An-Hung Hsiao, Kai-Ten Feng

In recent years, there have been abundant researches focused on indoor human presence detection based on laborious supervised learning (SL) and channel state information (CSI). These existing studies adopt spatial information of CSI to improve detection accuracy. However, channel is susceptible to arbitrary environmental changes in practice, such as the object movement, atmospheric factors and machine rebooting, which leads to degraded prediction accuracy. However, the existing SL-based methods require to re-train a new model with time-consuming labeling. Therefore, designing a semi-supervised learning (SSL) based scheme by continuously monitoring model "life-cycle" becomes compellingly imperative. In this paper, we propose bifold teacher-student (BTS) learning for presence detection system, which combines SSL by utilizing partial labeled and unlabeled dataset. The proposed primal-dual teacher-student network is capable of intelligently learning spatial and temporal features from labeled and unlabeled CSI. Additionally, the enhanced penalized loss function leveraging entropy and distance measure can distinguish the drifted data, i.e., features of new dataset are affected by time-varying effect and are alternated from the original distribution. The experimental results demonstrate that the proposed BTS system can sustain the asymptotic accuracy after retraining the model with unlabeled data. Moreover, label-free BTS outperforms the existing SSL-based models in terms of the highest detection accuracy, while achieving the similar performance of SL-based methods.

☆ Multi-modal Molecule Structure-text Model for Text-based Retrieval and Editing

Shengchao Liu, Weili Nie, Chengpeng Wang, Jiarui Lu, Zhuoran Qiao, Ling Liu, Jian Tang, Chaowei Xiao, Anima Anandkumar

There is increasing adoption of artificial intelligence in drug discovery. However, existing works use machine learning to mainly utilize the chemical structures of molecules yet ignore the vast textual knowledge available in chemistry. Incorporating textual knowledge enables us to realize new drug design objectives, adapt to text-based instructions, and predict complex biological activities. We present a multi-modal molecule structure-text model, MoleculeSTM, by jointly learning molecule's chemical structures and textual descriptions via a contrastive learning strategy. To train MoleculeSTM, we construct the largest multi-modal dataset to date, namely PubChemSTM, with over 280K chemical structure-text pairs. To demonstrate the effectiveness and utility of MoleculeSTM, we design two challenging zero-shot tasks based on text instructions, including structure-text retrieval and molecule editing. MoleculeSTM possesses two main properties: open vocabulary and compositionality via natural language. In experiments, MoleculeSTM obtains the state-of-the-art generalization ability to novel biochemical concepts across various benchmarks.

☆ GraphIX: Graph-based In silico XAI(explainable artificial intelligence) for drug repositioning from biopharmaceutical network

Atsuko Takagi, Mayumi Kamada, Eri Hamatani, Ryosuke Kojima, Yasushi Okuno

Drug repositioning holds great promise because it can reduce the time and cost of new drug development. While drug repositioning can omit various R&D processes, confirming pharmacological effects on biomolecules is essential for application to new diseases. Biomedical explainability in a drug repositioning model can support appropriate insights in subsequent in-depth studies. However, the validity of the XAI methodology is still under debate, and the effectiveness of XAI in drug repositioning prediction applications remains unclear. In this study, we propose GraphIX, an explainable drug repositioning framework using biological networks, and quantitatively evaluate its explainability. GraphIX first learns the network weights and node features using a graph neural network from known drug indication and knowledge graph that consists of three types of nodes (but not given node type information): disease, drug, and protein. Analysis of the post-learning features showed that node types that were not known to the model beforehand are distinguished through the learning process based on the graph structure. From the learned weights and features, GraphIX then predicts the disease-drug association and calculates the contribution values of the nodes located in the neighborhood of the predicted disease and drug. We hypothesized that the neighboring protein node to which the model gave a high contribution is important in understanding the actual pharmacological effects. Quantitative evaluation of the validity of protein nodes' contribution using a real-world database showed that the high contribution proteins shown by GraphIX are reasonable as a mechanism of drug action. GraphIX is a framework for evidence-based drug discovery that can present to users new disease-drug associations and identify the protein important for understanding its pharmacological effects from a large and complex knowledge base.

comment: 16 pages, 6 figures

☆ Multi-hop Evidence Retrieval for Cross-document Relation Extraction

Keming Lu, I-Hung Hsu, Wenxuan Zhou, Mingyu Derek Ma, Muhao Chen

Relation Extraction (RE) has been extended to cross-document scenarios because many relations are not simply described in a single document. This inevitably brings the challenge of efficient open-space evidence retrieval to support the inference of cross-document relations, along with the challenge of multi-hop reasoning on top of entities and evidence scattered in an open set of documents. To combat these challenges, we propose Mr.CoD, a multi-hop evidence retrieval method based on evidence path mining and ranking with adapted dense retrievers. We explore multiple variants of retrievers to show evidence retrieval is an essential part in cross-document RE. Experiments on CodRED show that evidence retrieval with Mr.Cod effectively acquires cross-document evidence that essentially supports open-setting cross-document RE. Additionally, we show that Mr.CoD facilitates evidence retrieval and boosts end-to-end RE performance with effective multi-hop reasoning in both closed and open settings of RE.

comment: Work in progress

☆ Hierarchically branched diffusion models for efficient and interpretable multi-class conditional generation

Alex M. Tseng, Tommaso Biancalani, Max Shen, Gabriele Scalia

Diffusion models have achieved justifiable popularity by attaining state-of-the-art performance in generating realistic objects from seemingly arbitrarily complex data distributions, including when conditioning generation on labels. Unfortunately, however, their iterative nature renders them very computationally inefficient during the sampling process. For the multi-class conditional generation problem, we propose a novel, structurally unique framework of diffusion models which are hierarchically branched according to the inherent relationships between classes. In this work, we demonstrate that branched diffusion models offer major improvements in efficiently generating samples from multiple classes. We also showcase several other advantages of branched diffusion models, including ease of extension to novel classes in a continual-learning setting, and a unique interpretability that offers insight into these generative models. Branched diffusion models represent an alternative paradigm to their traditional linear counterparts, and can have large impacts in how we use diffusion models for efficient generation, online learning, and scientific discovery.

☆ Towards Efficient Visual Simplification of Computational Graphs in Deep Neural Networks

Rusheng Pan, Zhiyong Wang, Yating Wei, Han Gao, Gongchang Ou, Caleb Chen Cao, Jingli Xu, Tong Xu, Wei Chen

A computational graph in a deep neural network (DNN) denotes a specific data flow diagram (DFD) composed of many tensors and operators. Existing toolkits for visualizing computational graphs are not applicable when the structure is highly complicated and large-scale (e.g., BERT [1]). To address this problem, we propose leveraging a suite of visual simplification techniques, including a cycle-removing method, a module-based edge-pruning algorithm, and an isomorphic subgraph stacking strategy. We design and implement an interactive visualization system that is suitable for computational graphs with up to 10 thousand elements. Experimental results and usage scenarios demonstrate that our tool reduces 60% elements on average and hence enhances the performance for recognizing and diagnosing DNN models. Our contributions are integrated into an open-source DNN visualization toolkit, namely, MindInsight [2].

☆ Reward Bonuses with Gain Scheduling Inspired by Iterative Deepening Search

Taisuke Kobayashi

This paper introduces a novel method of adding intrinsic bonuses to task-oriented reward function in order to efficiently facilitate reinforcement learning search. While various bonuses have been designed to date, they are analogous to the depth-first and breadth-first search algorithms in graph theory. This paper, therefore, first designs two bonuses for each of them. Then, a heuristic gain scheduling is applied to the designed bonuses, inspired by the iterative deepening search, which is known to inherit the advantages of the two search algorithms. The proposed method is expected to allow agent to efficiently reach the best solution in deeper states by gradually exploring unknown states. In three locomotion tasks with dense rewards and three simple tasks with sparse rewards, it is shown that the two types of bonuses contribute to the performance improvement of the different tasks complementarily. In addition, by combining them with the proposed gain scheduling, all tasks can be accomplished with high performance.

comment: 10 pages, 7 figures

☆ Learning List-Level Domain-Invariant Representations for Ranking

Ruicheng Xian, Honglei Zhuang, Zhen Qin, Hamed Zamani, Jing Lu, Ji Ma, Kai Hui, Han Zhao, Xuanhui Wang, Michael Bendersky

Domain adaptation aims to transfer the knowledge acquired by models trained on (data-rich) source domains to (low-resource) target domains, for which a popular method is invariant representation learning. While they have been studied extensively for classification and regression problems, how they apply to ranking problems, where the data and metrics have a list structure, is not well understood. Theoretically, we establish a domain adaptation generalization bound for ranking under listwise metrics such as MRR and NDCG. The bound suggests an adaptation method via learning list-level domain-invariant feature representations, whose benefits are empirically demonstrated by unsupervised domain adaptation experiments on real-world ranking tasks, including passage reranking. A key message is that for domain adaptation, the representations should be analyzed at the same level at which the metric is computed, as we show that learning invariant representations at the list level is most effective for adaptation on ranking problems.

☆ NADBenchmarks -- a compilation of Benchmark Datasets for Machine Learning Tasks related to Natural Disasters

Adiba Mahbub Proma, Md Saiful Islam, Stela Ciko, Raiyan Abdul Baten, Ehsan Hoque

Climate change has increased the intensity, frequency, and duration of extreme weather events and natural disasters across the world. While the increased data on natural disasters improves the scope of machine learning (ML) in this field, progress is relatively slow. One bottleneck is the lack of benchmark datasets that would allow ML researchers to quantify their progress against a standard metric. The objective of this short paper is to explore the state of benchmark datasets for ML tasks related to natural disasters, categorizing them according to the disaster management cycle. We compile a list of existing benchmark datasets introduced in the past five years. We propose a web platform - NADBenchmarks - where researchers can search for benchmark datasets for natural disasters, and we develop a preliminary version of such a platform using our compiled list. This paper is intended to aid researchers in finding benchmark datasets to train their ML models on, and provide general directions for topics where they can contribute new benchmark datasets.

☆ Scalable Hybrid Learning Techniques for Scientific Data Compression

Tania Banerjee, Jong Choi, Jaemoon Lee, Qian Gong, Jieyang Chen, Scott Klasky, Anand Rangarajan, Sanjay Ranka

Data compression is becoming critical for storing scientific data because many scientific applications need to store large amounts of data and post process this data for scientific discovery. Unlike image and video compression algorithms that limit errors to primary data, scientists require compression techniques that accurately preserve derived quantities of interest (QoIs). This paper presents a physics-informed compression technique implemented as an end-to-end, scalable, GPU-based pipeline for data compression that addresses this requirement. Our hybrid compression technique combines machine learning techniques and standard compression methods. Specifically, we combine an autoencoder, an error-bounded lossy compressor to provide guarantees on raw data error, and a constraint satisfaction post-processing step to preserve the QoIs within a minimal error (generally less than floating point error). The effectiveness of the data compression pipeline is demonstrated by compressing nuclear fusion simulation data generated by a large-scale fusion code, XGC, which produces hundreds of terabytes of data in a single day. Our approach works within the ADIOS framework and results in compression by a factor of more than 150 while requiring only a few percent of the computational resources necessary for generating the data, making the overall approach highly effective for practical scenarios.

☆ UnICLAM:Contrastive Representation Learning with Adversarial Masking for Unified and Interpretable Medical Vision Question Answering

Chenlu Zhan, Peng Peng, Hongsen Wang, Tao Chen, Hongwei Wang

Medical Visual Question Answering (Medical-VQA) aims to answer clinical questions regarding radiology images, assisting doctors with decision-making options. Nevertheless, current Medical-VQA models learn cross-modal representations through residing vision and texture encoders in dual separate spaces, which lead to indirect semantic alignment. In this paper, we propose UnICLAM, a Unified and Interpretable Medical-VQA model through Contrastive Representation Learning with Adversarial Masking. Specifically, to learn an aligned image-text representation, we first establish a unified dual-stream pre-training structure with the gradually soft-parameter sharing strategy. Technically, the proposed strategy learns a constraint for the vision and texture encoders to be close in a same space, which is gradually loosened as the higher number of layers. Moreover, for grasping the semantic representation, we extend the unified Adversarial Masking data augmentation strategy to the contrastive representation learning of vision and text in a unified manner, alleviating the meaningless of the commonly used random mask. Concretely, while the encoder training minimizes the distance between the original feature and the masking feature, the adversarial masking model keeps adversarial learning to conversely maximize the distance. Furthermore, we also intuitively take a further exploration of the unified adversarial masking strategy, which improves the potential ante-hoc interpretability with remarkable performance and efficiency. Experimental results on VQA-RAD and SLAKE public benchmarks demonstrate that UnICLAM outperforms the existing 11 state-of-the-art Medical-VQA models. More importantly, we make an additional discussion about the performance of UnICLAM in diagnosing heart failure, verifying that UnICLAM exhibits superior few-shot adaption performance in practical disease diagnosis.

☆ Beyond Contrastive Learning: A Variational Generative Model for Multilingual Retrieval

John Wieting, Jonathan H. Clark, William W. Cohen, Graham Neubig, Taylor Berg-Kirkpatrick

Contrastive learning has been successfully used for retrieval of semantically aligned sentences, but it often requires large batch sizes or careful engineering to work well. In this paper, we instead propose a generative model for learning multilingual text embeddings which can be used to retrieve or score sentence pairs. Our model operates on parallel data in $N$ languages and, through an approximation we introduce, efficiently encourages source separation in this multilingual setting, separating semantic information that is shared between translations from stylistic or language-specific variation. We show careful large-scale comparisons between contrastive and generation-based approaches for learning multilingual text embeddings, a comparison that has not been done to the best of our knowledge despite the popularity of these approaches. We evaluate this method on a suite of tasks including semantic similarity, bitext mining, and cross-lingual question retrieval -- the last of which we introduce in this paper. Overall, our Variational Multilingual Source-Separation Transformer (VMSST) model outperforms both a strong contrastive and generative baseline on these tasks.

☆ Interpretability and causal discovery of the machine learning models to predict the production of CBM wells after hydraulic fracturing

Chao Min, Guoquan Wen, Liangjie Gou, Xiaogang Li, Zhaozhong Yang

Machine learning approaches are widely studied in the production prediction of CBM wells after hydraulic fracturing, but merely used in practice due to the low generalization ability and the lack of interpretability. A novel methodology is proposed in this article to discover the latent causality from observed data, which is aimed at finding an indirect way to interpret the machine learning results. Based on the theory of causal discovery, a causal graph is derived with explicit input, output, treatment and confounding variables. Then, SHAP is employed to analyze the influence of the factors on the production capability, which indirectly interprets the machine learning models. The proposed method can capture the underlying nonlinear relationship between the factors and the output, which remedies the limitation of the traditional machine learning routines based on the correlation analysis of factors. The experiment on the data of CBM shows that the detected relationship between the production and the geological/engineering factors by the presented method, is coincident with the actual physical mechanism. Meanwhile, compared with traditional methods, the interpretable machine learning models have better performance in forecasting production capability, averaging 20% improvement in accuracy.

☆ Hidden Poison: Machine Unlearning Enables Camouflaged Poisoning Attacks

Jimmy Z. Di, Jack Douglas, Jayadev Acharya, Gautam Kamath, Ayush Sekhari

We introduce camouflaged data poisoning attacks, a new attack vector that arises in the context of machine unlearning and other settings when model retraining may be induced. An adversary first adds a few carefully crafted points to the training dataset such that the impact on the model's predictions is minimal. The adversary subsequently triggers a request to remove a subset of the introduced points at which point the attack is unleashed and the model's predictions are negatively affected. In particular, we consider clean-label targeted attacks (in which the goal is to cause the model to misclassify a specific test point) on datasets including CIFAR-10, Imagenette, and Imagewoof. This attack is realized by constructing camouflage datapoints that mask the effect of a poisoned dataset.

☆ Neighboring state-based RL Exploration

Jeffery Cheng, Kevin Li, Justin Lin, Pedro Pachuca

Reinforcement Learning is a powerful tool to model decision-making processes. However, it relies on an exploration-exploitation trade-off that remains an open challenge for many tasks. In this work, we study neighboring state-based, model-free exploration led by the intuition that, for an early-stage agent, considering actions derived from a bounded region of nearby states may lead to better actions when exploring. We propose two algorithms that choose exploratory actions based on a survey of nearby states, and find that one of our methods, ${\rho}$-explore, consistently outperforms the Double DQN baseline in an discrete environment by 49\% in terms of Eval Reward Return.

☆ Extractive Text Summarization Using Generalized Additive Models with Interactions for Sentence Selection

Vinícius Camargo da Silva, João Paulo Papa, Kelton Augusto Pontara da Costa

Automatic Text Summarization (ATS) is becoming relevant with the growth of textual data; however, with the popularization of public large-scale datasets, some recent machine learning approaches have focused on dense models and architectures that, despite producing notable results, usually turn out in models difficult to interpret. Given the challenge behind interpretable learning-based text summarization and the importance it may have for evolving the current state of the ATS field, this work studies the application of two modern Generalized Additive Models with interactions, namely Explainable Boosting Machine and GAMI-Net, to the extractive summarization problem based on linguistic features and binary classification.

☆ Control of Continuous Quantum Systems with Many Degrees of Freedom based on Convergent Reinforcement Learning

Zhikang Wang

With the development of experimental quantum technology, quantum control has attracted increasing attention due to the realization of controllable artificial quantum systems. However, because quantum-mechanical systems are often too difficult to analytically deal with, heuristic strategies and numerical algorithms which search for proper control protocols are adopted, and, deep learning, especially deep reinforcement learning (RL), is a promising generic candidate solution for the control problems. Although there have been a few successful applications of deep RL to quantum control problems, most of the existing RL algorithms suffer from instabilities and unsatisfactory reproducibility, and require a large amount of fine-tuning and a large computational budget, both of which limit their applicability. To resolve the issue of instabilities, in this dissertation, we investigate the non-convergence issue of Q-learning. Then, we investigate the weakness of existing convergent approaches that have been proposed, and we develop a new convergent Q-learning algorithm, which we call the convergent deep Q network (C-DQN) algorithm, as an alternative to the conventional deep Q network (DQN) algorithm. We prove the convergence of C-DQN and apply it to the Atari 2600 benchmark. We show that when DQN fail, C-DQN still learns successfully. Then, we apply the algorithm to the measurement-feedback cooling problems of a quantum quartic oscillator and a trapped quantum rigid body. We establish the physical models and analyse their properties, and we show that although both C-DQN and DQN can learn to cool the systems, C-DQN tends to behave more stably, and when DQN suffers from instabilities, C-DQN can achieve a better performance. As the performance of DQN can have a large variance and lack consistency, C-DQN can be a better choice for researches on complicated control problems.

comment: PhD Dissertation submitted to the Department of Physics, University of Tokyo

☆ A Non-Asymptotic Analysis of Oversmoothing in Graph Neural Networks

Xinyi Wu, Zhengdao Chen, William Wang, Ali Jadbabaie

A central challenge of building more powerful Graph Neural Networks (GNNs) is the oversmoothing phenomenon, where increasing the network depth leads to homogeneous node representations and thus worse classification performance. While previous works have only demonstrated that oversmoothing is inevitable when the number of graph convolutions tends to infinity, in this paper, we precisely characterize the mechanism behind the phenomenon via a non-asymptotic analysis. Specifically, we distinguish between two different effects when applying graph convolutions -- an undesirable mixing effect that homogenizes node representations in different classes, and a desirable denoising effect that homogenizes node representations in the same class. By quantifying these two effects on random graphs sampled from the Contextual Stochastic Block Model (CSBM), we show that oversmoothing happens once the mixing effect starts to dominate the denoising effect, and the number of layers required for this transition is $O(\log N/\log (\log N))$ for sufficiently dense graphs with $N$ nodes. We also extend our analysis to study the effects of Personalized PageRank (PPR) on oversmoothing. Our results suggest that while PPR mitigates oversmoothing at deeper layers, PPR-based architectures still achieve their best performance at a shallow depth and are outperformed by the graph convolution approach on certain graphs. Finally, we support our theoretical results with numerical experiments, which further suggest that the oversmoothing phenomenon observed in practice may be exacerbated by the difficulty of optimizing deep GNN models.

☆ Analyzing Semantic Faithfulness of Language Models via Input Intervention on Conversational Question Answering

Akshay Chaturvedi, Swarnadeep Bhar, Soumadeep Saha, Utpal Garain, Nicholas Asher

Transformer-based language models have been shown to be highly effective for several NLP tasks. In this paper, we consider three transformer models, BERT, RoBERTa, and XLNet, in both small and large version, and investigate how faithful their representations are with respect to the semantic content of texts. We formalize a notion of semantic faithfulness, in which the semantic content of a text should causally figure in a model's inferences in question answering. We then test this notion by observing a model's behavior on answering questions about a story after performing two novel semantic interventions -- deletion intervention and negation intervention. While transformer models achieve high performance on standard question answering tasks, we show that they fail to be semantically faithful once we perform these interventions for a significant number of cases (~50% for deletion intervention, and ~20% drop in accuracy for negation intervention). We then propose an intervention-based training regime that can mitigate the undesirable effects for deletion intervention by a significant margin (from ~50% to ~6%). We analyze the inner-workings of the models to better understand the effectiveness of intervention-based training for deletion intervention. But we show that this training does not attenuate other aspects of semantic unfaithfulness such as the models' inability to deal with negation intervention or to capture the predicate-argument structure of texts. We also test InstructGPT, via prompting, for its ability to handle the two interventions and to capture predicate-argument structure. While InstructGPT models do achieve very high performance on predicate-argument structure task, they fail to respond adequately to our deletion and negation interventions.

comment: 27 pages, 4 figures

☆ Circumventing interpretability: How to defeat mind-readers

Lee Sharkey

The increasing capabilities of artificial intelligence (AI) systems make it ever more important that we interpret their internals to ensure that their intentions are aligned with human values. Yet there is reason to believe that misaligned artificial intelligence will have a convergent instrumental incentive to make its thoughts difficult for us to interpret. In this article, I discuss many ways that a capable AI might circumvent scalable interpretability methods and suggest a framework for thinking about these potential future risks.

☆ Improving Automated Program Repair with Domain Adaptation

Armin Zirak, Hadi Hemati

Automated Program Repair (APR) is defined as the process of fixing a bug/defect in the source code, by an automated tool. APR tools have recently experienced promising results by leveraging state-of-the-art Neural Language Processing (NLP) techniques. APR tools such as TFix and CodeXGLUE combine text-to-text transformers with software-specific techniques are outperforming alternatives, these days. However, in most APR studies the train and test sets are chosen from the same set of projects. In reality, however, APR models are meant to be generalizable to new and different projects. Therefore, there is a potential threat that reported APR models with high effectiveness perform poorly when the characteristics of the new project or its bugs are different than the training set's(Domain Shift). In this study, we first define and measure the domain shift problem in automated program repair. Then, we then propose a domain adaptation framework that can adapt an APR model for a given target project. We conduct an empirical study with three domain adaptation methods FullFineTuning, TuningWithLightWeightAdapterLayers, and CurriculumLearning using two state-of-the-art domain adaptation tools (TFix and CodeXGLUE) and two APR models on 611 bugs from 19 projects. The results show that our proposed framework can improve the effectiveness of TFix by 13.05% and CodeXGLUE by 23.4%. Another contribution of this study is the proposal of a data synthesis method to address the lack of labelled data in APR. We leverage transformers to create a bug generator model. We use the generated synthetic data to domain adapt TFix and CodeXGLUE on the projects with no data (Zero-shot learning), which results in an average improvement of 5.76% and 24.42% for TFix and CodeXGLUE, respectively.

comment: 43 pages

☆ DExT: Detector Explanation Toolkit

Deepan Chakravarthi Padmanabhan, Matias Valdenegro-Toro

State-of-the-art object detectors are treated as black boxes due to their highly non-linear internal computations. Even with unprecedented advancements in detector performance, the inability to explain how their outputs are generated limits their use in safety-critical applications. Previous work fails to produce explanations for both bounding box and classification decisions, and generally make individual explanations for various detectors. In this paper, we propose an open-source Detector Explanation Toolkit (DExT) which implements the proposed approach to generate a holistic explanation for all detector decisions using certain gradient-based explanation methods. We suggests various multi-object visualization methods to merge the explanations of multiple objects detected in an image as well as the corresponding detections in a single image. The quantitative evaluation show that the Single Shot MultiBox Detector (SSD) is more faithfully explained compared to other detectors regardless of the explanation methods. Both quantitative and human-centric evaluations identify that SmoothGrad with Guided Backpropagation (GBP) provides more trustworthy explanations among selected methods across all detectors. We expect that DExT will motivate practitioners to evaluate object detectors from the interpretability perspective by explaining both bounding box and classification decisions.

comment: 21 pages, with supplementary

☆ Adaptive and Dynamic Multi-Resolution Hashing for Pairwise Summations

Lianke Qin, Aravind Reddy, Zhao Song, Zhaozhuo Xu, Danyang Zhuo

In this paper, we propose Adam-Hash: an adaptive and dynamic multi-resolution hashing data-structure for fast pairwise summation estimation. Given a data-set $X \subset \mathbb{R}^d$, a binary function $f:\mathbb{R}^d\times \mathbb{R}^d\to \mathbb{R}$, and a point $y \in \mathbb{R}^d$, the Pairwise Summation Estimate $\mathrm{PSE}_X(y) := \frac{1}{|X|} \sum_{x \in X} f(x,y)$. For any given data-set $X$, we need to design a data-structure such that given any query point $y \in \mathbb{R}^d$, the data-structure approximately estimates $\mathrm{PSE}_X(y)$ in time that is sub-linear in $|X|$. Prior works on this problem have focused exclusively on the case where the data-set is static, and the queries are independent. In this paper, we design a hashing-based PSE data-structure which works for the more practical \textit{dynamic} setting in which insertions, deletions, and replacements of points are allowed. Moreover, our proposed Adam-Hash is also robust to adaptive PSE queries, where an adversary can choose query $q_j \in \mathbb{R}^d$ depending on the output from previous queries $q_1, q_2, \dots, q_{j-1}$.

comment: BigData 2022

☆ ABODE-Net: An Attention-based Deep Learning Model for Non-intrusive Building Occupancy Detection Using Smart Meter Data

Zhirui Luo, Ruobin Qi, Qingqing Li, Jun Zheng, Sihua Shao

Occupancy information is useful for efficient energy management in the building sector. The massive high-resolution electrical power consumption data collected by smart meters in the advanced metering infrastructure (AMI) network make it possible to infer buildings' occupancy status in a non-intrusive way. In this paper, we propose a deep leaning model called ABODE-Net which employs a novel Parallel Attention (PA) block for building occupancy detection using smart meter data. The PA block combines the temporal, variable, and channel attention modules in a parallel way to signify important features for occupancy detection. We adopt two smart meter datasets widely used for building occupancy detection in our performance evaluation. A set of state-of-the-art shallow machine learning and deep learning models are included for performance comparison. The results show that ABODE-Net significantly outperforms other models in all experimental cases, which proves its validity as a solution for non-intrusive building occupancy detection.

comment: To be published in The 7th International Conference on Smart Computing and Communication (SmartCom 2022)

☆ Online Statistical Inference for Matrix Contextual Bandit

Qiyu Han, Will Wei Sun, Yichen Zhang

Contextual bandit has been widely used for sequential decision-making based on the current contextual information and historical feedback data. In modern applications, such context format can be rich and can often be formulated as a matrix. Moreover, while existing bandit algorithms mainly focused on reward-maximization, less attention has been paid to the statistical inference. To fill in these gaps, in this work we consider a matrix contextual bandit framework where the true model parameter is a low-rank matrix, and propose a fully online procedure to simultaneously make sequential decision-making and conduct statistical inference. The low-rank structure of the model parameter and the adaptivity nature of the data collection process makes this difficult: standard low-rank estimators are not fully online and are biased, while existing inference approaches in bandit algorithms fail to account for the low-rankness and are also biased. To address these, we introduce a new online doubly-debiasing inference procedure to simultaneously handle both sources of bias. In theory, we establish the asymptotic normality of the proposed online doubly-debiased estimator and prove the validity of the constructed confidence interval. Our inference results are built upon a newly developed low-rank stochastic gradient descent estimator and its non-asymptotic convergence result, which is also of independent interest.

comment: 81 pages

☆ ReVISE: Self-Supervised Speech Resynthesis with Visual Input for Universal and Generalized Speech Enhancement

Wei-Ning Hsu, Tal Remez, Bowen Shi, Jacob Donley, Yossi Adi

Prior works on improving speech quality with visual input typically study each type of auditory distortion separately (e.g., separation, inpainting, video-to-speech) and present tailored algorithms. This paper proposes to unify these subjects and study Generalized Speech Enhancement, where the goal is not to reconstruct the exact reference clean signal, but to focus on improving certain aspects of speech. In particular, this paper concerns intelligibility, quality, and video synchronization. We cast the problem as audio-visual speech resynthesis, which is composed of two steps: pseudo audio-visual speech recognition (P-AVSR) and pseudo text-to-speech synthesis (P-TTS). P-AVSR and P-TTS are connected by discrete units derived from a self-supervised speech model. Moreover, we utilize self-supervised audio-visual speech model to initialize P-AVSR. The proposed model is coined ReVISE. ReVISE is the first high-quality model for in-the-wild video-to-speech synthesis and achieves superior performance on all LRS3 audio-visual enhancement tasks with a single model. To demonstrates its applicability in the real world, ReVISE is also evaluated on EasyCom, an audio-visual benchmark collected under challenging acoustic conditions with only 1.6 hours of training data. Similarly, ReVISE greatly suppresses noise and improves quality. Project page: https://wnhsu.github.io/ReVISE.

☆ Semi-supervised GAN for Bladder Tissue Classification in Multi-Domain Endoscopic Images

Jorge F. Lazo, Benoit Rosa, Michele Catellani, Matteo Fontana, Francesco A. Mistretta, Gennaro Musi, Ottavio de Cobelli, Michel de Mathelin, Elena De Momi

Objective: Accurate visual classification of bladder tissue during Trans-Urethral Resection of Bladder Tumor (TURBT) procedures is essential to improve early cancer diagnosis and treatment. During TURBT interventions, White Light Imaging (WLI) and Narrow Band Imaging (NBI) techniques are used for lesion detection. Each imaging technique provides diverse visual information that allows clinicians to identify and classify cancerous lesions. Computer vision methods that use both imaging techniques could improve endoscopic diagnosis. We address the challenge of tissue classification when annotations are available only in one domain, in our case WLI, and the endoscopic images correspond to an unpaired dataset, i.e. there is no exact equivalent for every image in both NBI and WLI domains. Method: We propose a semi-surprised Generative Adversarial Network (GAN)-based method composed of three main components: a teacher network trained on the labeled WLI data; a cycle-consistency GAN to perform unpaired image-to-image translation, and a multi-input student network. To ensure the quality of the synthetic images generated by the proposed GAN we perform a detailed quantitative, and qualitative analysis with the help of specialists. Conclusion: The overall average classification accuracy, precision, and recall obtained with the proposed method for tissue classification are 0.90, 0.88, and 0.89 respectively, while the same metrics obtained in the unlabeled domain (NBI) are 0.92, 0.64, and 0.94 respectively. The quality of the generated images is reliable enough to deceive specialists. Significance: This study shows the potential of using semi-supervised GAN-based classification to improve bladder tissue classification when annotations are limited in multi-domain data.

☆ MM811 Project Report: Cloud Detection and Removal in Satellite Images

Dale Chen-Song, Erfan Khalaji, Vaishali Rani

For satellite images, the presence of clouds presents a problem as clouds obscure more than half to two-thirds of the ground information. This problem causes many issues for reliability in a noise-free environment to communicate data and other applications that need seamless monitoring. Removing the clouds from the images while keeping the background pixels intact can help address the mentioned issues. Recently, deep learning methods have become popular for researching cloud removal by demonstrating promising results, among which Generative Adversarial Networks (GAN) have shown considerably better performance. In this project, we aim to address cloud removal from satellite images using AttentionGAN and then compare our results by reproducing the results obtained using traditional GANs and auto-encoders. We use RICE dataset. The outcome of this project can be used to develop applications that require cloud-free satellite images. Moreover, our results could be helpful for making further research improvements.

☆ When and Why Test Generators for Deep Learning Produce Invalid Inputs: an Empirical Study ICSE 2023

Vincenzo Riccio, Paolo Tonella

Testing Deep Learning (DL) based systems inherently requires large and representative test sets to evaluate whether DL systems generalise beyond their training datasets. Diverse Test Input Generators (TIGs) have been proposed to produce artificial inputs that expose issues of the DL systems by triggering misbehaviours. Unfortunately, such generated inputs may be invalid, i.e., not recognisable as part of the input domain, thus providing an unreliable quality assessment. Automated validators can ease the burden of manually checking the validity of inputs for human testers, although input validity is a concept difficult to formalise and, thus, automate. In this paper, we investigate to what extent TIGs can generate valid inputs, according to both automated and human validators. We conduct a large empirical study, involving 2 different automated validators, 220 human assessors, 5 different TIGs and 3 classification tasks. Our results show that 84% artificially generated inputs are valid, according to automated validators, but their expected label is not always preserved. Automated validators reach a good consensus with humans (78% accuracy), but still have limitations when dealing with feature-rich datasets.

comment: To be published in Proceedings of the 45th ACM/IEEE International Conference on Software Engineering (ICSE 2023)

☆ Forecasting West Nile Virus with Graph Neural Networks: Harnessing Spatial Dependence in Irregularly Sampled Geospatial Data

Adam Tonks, Trevor Harris, Bo Li, William Brown, Rebecca Smith

Machine learning methods have seen increased application to geospatial environmental problems, such as precipitation nowcasting, haze forecasting, and crop yield prediction. However, many of the machine learning methods applied to mosquito population and disease forecasting do not inherently take into account the underlying spatial structure of the given data. In our work, we apply a spatially aware graph neural network model consisting of GraphSAGE layers to forecast the presence of West Nile virus in Illinois, to aid mosquito surveillance and abatement efforts within the state. More generally, we show that graph neural networks applied to irregularly sampled geospatial data can exceed the performance of a range of baseline methods including logistic regression, XGBoost, and fully-connected neural networks.

☆ Feature Acquisition using Monte Carlo Tree Search

Sungsoo Lim, Diego Klabjan, Mark Shapiro

Feature acquisition algorithms address the problem of acquiring informative features while balancing the costs of acquisition to improve the learning performances of ML models. Previous approaches have focused on calculating the expected utility values of features to determine the acquisition sequences. Other approaches formulated the problem as a Markov Decision Process (MDP) and applied reinforcement learning based algorithms. In comparison to previous approaches, we focus on 1) formulating the feature acquisition problem as a MDP and applying Monte Carlo Tree Search, 2) calculating the intermediary rewards for each acquisition step based on model improvements and acquisition costs and 3) simultaneously optimizing model improvement and acquisition costs with multi-objective Monte Carlo Tree Search. With Proximal Policy Optimization and Deep Q-Network algorithms as benchmark, we show the effectiveness of our proposed approach with experimental study.

comment: 13 pages, 7 figures

☆ Contrastive Distillation Is a Sample-Efficient Self-Supervised Loss Policy for Transfer Learning

Chris Lengerich, Gabriel Synnaeve, Amy Zhang, Hugh Leather, Kurt Shuster, François Charton, Charysse Redwood

Traditional approaches to RL have focused on learning decision policies directly from episodic decisions, while slowly and implicitly learning the semantics of compositional representations needed for generalization. While some approaches have been adopted to refine representations via auxiliary self-supervised losses while simultaneously learning decision policies, learning compositional representations from hand-designed and context-independent self-supervised losses (multi-view) still adapts relatively slowly to the real world, which contains many non-IID subspaces requiring rapid distribution shift in both time and spatial attention patterns at varying levels of abstraction. In contrast, supervised language model cascades have shown the flexibility to adapt to many diverse manifolds, and hints of self-learning needed for autonomous task transfer. However, to date, transfer methods for language models like few-shot learning and fine-tuning still require human supervision and transfer learning using self-learning methods has been underexplored. We propose a self-supervised loss policy called contrastive distillation which manifests latent variables with high mutual information with both source and target tasks from weights to tokens. We show how this outperforms common methods of transfer learning and suggests a useful design axis of trading off compute for generalizability for online transfer. Contrastive distillation is improved through sampling from memory and suggests a simple algorithm for more efficiently sampling negative examples for contrastive losses than random sampling.

☆ Sensitivity analysis of biological washout and depth selection for a machine learning based dose verification framework in proton therapy

Shixiong Yu, Yuxiang Liu, Zongsheng Hu, Haozhao Zhang, Pengyu Qi, Hao Peng

Dose verification based on proton-induced positron emitters is a promising quality assurance tool and may leverage the strength of artificial intelligence. To move a step closer towards practical application, the sensitivity analysis of two factors needs to be performed: biological washout and depth selection. selection. A bi-directional recurrent neural network (RNN) model was developed. The training dataset was generated based upon a CT image-based phantom (abdomen region) and multiple beam energies/pathways, using Monte-Carlo simulation (1 mm spatial resolution, no biological washout). For the modeling of biological washout, a simplified analytical model was applied to change raw activity profiles over a period of 5 minutes, incorporating both physical decay and biological washout. For the study of depth selection (a challenge linked to multi field/angle irradiation), truncations were applied at different window lengths (100, 125, 150 mm) to raw activity profiles. Finally, the performance of a worst-case scenario was examined by combining both factors (depth selection: 125 mm, biological washout: 5 mins). The accuracy was quantitatively evaluated in terms of range uncertainty, mean absolute error (MAE) and mean relative errors (MRE). Our proposed AI framework shows good immunity to the perturbation associated with two factors. The detection of proton-induced positron emitters, combined with machine learning, has great potential to implement online patient-specific verification in proton therapy.

☆ Deep Unfolded Tensor Robust PCA with Self-supervised Learning

Harry Dong, Megna Shah, Sean Donegan, Yuejie Chi

Tensor robust principal component analysis (RPCA), which seeks to separate a low-rank tensor from its sparse corruptions, has been crucial in data science and machine learning where tensor structures are becoming more prevalent. While powerful, existing tensor RPCA algorithms can be difficult to use in practice, as their performance can be sensitive to the choice of additional hyperparameters, which are not straightforward to tune. In this paper, we describe a fast and simple self-supervised model for tensor RPCA using deep unfolding by only learning four hyperparameters. Despite its simplicity, our model expunges the need for ground truth labels while maintaining competitive or even greater performance compared to supervised deep unfolding. Furthermore, our model is capable of operating in extreme data-starved scenarios. We demonstrate these claims on a mix of synthetic data and real-world tasks, comparing performance against previously studied supervised deep unfolding methods and Bayesian optimization baselines.

☆ Target Conditioned Representation Independence (TCRI); From Domain-Invariant to Domain-General Representations

Olawale Salaudeen, Oluwasanmi Koyejo

We propose a Target Conditioned Representation Independence (TCRI) objective for domain generalization. TCRI addresses the limitations of existing domain generalization methods due to incomplete constraints. Specifically, TCRI implements regularizers motivated by conditional independence constraints that are sufficient to strictly learn complete sets of invariant mechanisms, which we show are necessary and sufficient for domain generalization. Empirically, we show that TCRI is effective on both synthetic and real-world data. TCRI is competitive with baselines in average accuracy while outperforming them in worst-domain accuracy, indicating desired cross-domain stability.

☆ Debiased machine learning for estimating the causal effect of urban traffic on pedestrian crossing behaviour

Kimia Kamal, Bilal Farooq

Before the transition of AVs to urban roads and subsequently unprecedented changes in traffic conditions, evaluation of transportation policies and futuristic road design related to pedestrian crossing behavior is of vital importance. Recent studies analyzed the non-causal impact of various variables on pedestrian waiting time in the presence of AVs. However, we mainly investigate the causal effect of traffic density on pedestrian waiting time. We develop a Double/Debiased Machine Learning (DML) model in which the impact of confounders variable influencing both a policy and an outcome of interest is addressed, resulting in unbiased policy evaluation. Furthermore, we try to analyze the effect of traffic density by developing a copula-based joint model of two main components of pedestrian crossing behavior, pedestrian stress level and waiting time. The copula approach has been widely used in the literature, for addressing self-selection problems, which can be classified as a causality analysis in travel behavior modeling. The results obtained from copula approach and DML are compared based on the effect of traffic density. In DML model structure, the standard error term of density parameter is lower than copula approach and the confidence interval is considerably more reliable. In addition, despite the similar sign of effect, the copula approach estimates the effect of traffic density lower than DML, due to the spurious effect of confounders. In short, the DML model structure can flexibly adjust the impact of confounders by using machine learning algorithms and is more reliable for planning future policies.

☆ End-to-end AI Framework for Hyperparameter Optimization, Model Training, and Interpretable Inference for Molecules and Crystals

Hyun Park, Ruijie Zhu, E. A. Huerta, Santanu Chaudhuri, Emad Tajkhorshid, Donny Cooper

We introduce an end-to-end computational framework that enables hyperparameter optimization with the DeepHyper library, accelerated training, and interpretable AI inference with a suite of state-of-the-art AI models, including CGCNN, PhysNet, SchNet, MPNN, MPNN-transformer, and TorchMD-Net. We use these AI models and the benchmark QM9, hMOF, and MD17 datasets to showcase the prediction of user-specified materials properties in modern computing environments, and to demonstrate translational applications for the modeling of small molecules, crystals and metal organic frameworks with a unified, stand-alone framework. We deployed and tested this framework in the ThetaGPU supercomputer at the Argonne Leadership Computing Facility, and the Delta supercomputer at the National Center for Supercomputing Applications to provide researchers with modern tools to conduct accelerated AI-driven discovery in leadership class computing environments.

comment: 20 pages, 10 images, 6 tables

☆ What do LLMs Know about Financial Markets? A Case Study on Reddit Market Sentiment Analysis

Xiang Deng, Vasilisa Bashlovkina, Feng Han, Simon Baumgartner, Michael Bendersky

Market sentiment analysis on social media content requires knowledge of both financial markets and social media jargon, which makes it a challenging task for human raters. The resulting lack of high-quality labeled data stands in the way of conventional supervised learning methods. Instead, we approach this problem using semi-supervised learning with a large language model (LLM). Our pipeline generates weak financial sentiment labels for Reddit posts with an LLM and then uses that data to train a small model that can be served in production. We find that prompting the LLM to produce Chain-of-Thought summaries and forcing it through several reasoning paths helps generate more stable and accurate labels, while using a regression loss further improves distillation quality. With only a handful of prompts, the final model performs on par with existing supervised models. Though production applications of our model are limited by ethical considerations, the model's competitive performance points to the great potential of using LLMs for tasks that otherwise require skill-intensive annotation.

☆ Towards Neural Variational Monte Carlo That Scales Linearly with System Size NeurIPS 2022

Or Sharir, Garnet Kin-Lic Chan, Anima Anandkumar

Quantum many-body problems are some of the most challenging problems in science and are central to demystifying some exotic quantum phenomena, e.g., high-temperature superconductors. The combination of neural networks (NN) for representing quantum states, coupled with the Variational Monte Carlo (VMC) algorithm, has been shown to be a promising method for solving such problems. However, the run-time of this approach scales quadratically with the number of simulated particles, constraining the practically usable NN to - in machine learning terms - minuscule sizes (<10M parameters). Considering the many breakthroughs brought by extreme NN in the +1B parameters scale to other domains, lifting this constraint could significantly expand the set of quantum systems we can accurately simulate on classical computers, both in size and complexity. We propose a NN architecture called Vector-Quantized Neural Quantum States (VQ-NQS) that utilizes vector-quantization techniques to leverage redundancies in the local-energy calculations of the VMC algorithm - the source of the quadratic scaling. In our preliminary experiments, we demonstrate VQ-NQS ability to reproduce the ground state of the 2D Heisenberg model across various system sizes, while reporting a significant reduction of about ${\times}10$ in the number of FLOPs in the local-energy calculation.

comment: Appeared on NeurIPS 2022 AI for Science Workshop (a non-archival poster presentation)

☆ Language models are better than humans at next-token prediction

Buck Shlegeris, Fabien Roger, Lawrence Chan, Euan McLean

Current language models are considered to have sub-human capabilities at natural language tasks like question-answering or writing code. However, language models are not trained to perform well at these tasks, they are trained to accurately predict the next token given previous tokes in tokenized text. It is not clear whether language models are better or worse than humans at next token prediction. To try to answer this question, we performed two distinct experiments to directly compare humans and language models on this front: one measuring top-1 accuracy and the other measuring perplexity. In both experiments, we find humans to be consistently \emph{worse} than even relatively small language models like GPT3-Ada at next-token prediction.

☆ Decision-making and control with metasurface-based diffractive neural networks

Jumin Qiu, Tianbao Yu, Lujun Huang, Andrey Miroshnichenko, Shuyuan Xiao

The ultimate goal of artificial intelligence is to mimic the human brain to perform decision-making and control directly from high-dimensional sensory input. All-optical diffractive neural networks provide a promising solution for realizing artificial intelligence with high-speed and low-power consumption. To date, most of the reported diffractive neural networks focus on single or multiple tasks that do not involve interaction with the environment, such as object recognition and image classification, while the networks that can perform decision-making and control, to our knowledge, have not been developed yet. Here, we propose to use deep reinforcement learning to realize diffractive neural networks that enable imitating the human-level capability of decision-making and control. Such networks allow for finding optimal control policies through interaction with the environment and can be readily realized with the dielectric metasurfaces. The superior performances of these networks are verified by engaging three types of classic games, Tic-Tac-Toe, Super Mario Bros., and Car Racing, and achieving the same or even higher levels comparable to human players. Our work represents a solid step of advancement in diffractive neural networks, which promises a fundamental shift from the target-driven control of a pre-designed state for simple recognition or classification tasks to the high-level sensory capability of artificial intelligence. It may find exciting applications in autonomous driving, intelligent robots, and intelligent manufacturing.

♻ ☆ A Tunable Loss Function for Robust Classification: Calibration, Landscape, and Generalization

Tyler Sypherd, Mario Diaz, John Kevin Cava, Gautam Dasarathy, Peter Kairouz, Lalitha Sankar

We introduce a tunable loss function called $\alpha$-loss, parameterized by $\alpha \in (0,\infty]$, which interpolates between the exponential loss ($\alpha = 1/2$), the log-loss ($\alpha = 1$), and the 0-1 loss ($\alpha = \infty$), for the machine learning setting of classification. Theoretically, we illustrate a fundamental connection between $\alpha$-loss and Arimoto conditional entropy, verify the classification-calibration of $\alpha$-loss in order to demonstrate asymptotic optimality via Rademacher complexity generalization techniques, and build-upon a notion called strictly local quasi-convexity in order to quantitatively characterize the optimization landscape of $\alpha$-loss. Practically, we perform class imbalance, robustness, and classification experiments on benchmark image datasets using convolutional-neural-networks. Our main practical conclusion is that certain tasks may benefit from tuning $\alpha$-loss away from log-loss ($\alpha = 1$), and to this end we provide simple heuristics for the practitioner. In particular, navigating the $\alpha$ hyperparameter can readily provide superior model robustness to label flips ($\alpha > 1$) and sensitivity to imbalanced classes ($\alpha < 1$).

comment: Published at the Transactions on Information Theory

♻ ☆ FAIR AI Models in High Energy Physics

Javier Duarte, Haoyang Li, Avik Roy, Ruike Zhu, E. A. Huerta, Daniel Diaz, Philip Harris, Raghav Kansal, Daniel S. Katz, Ishaan H. Kavoori, Volodymyr V. Kindratenko, Farouk Mokhtar, Mark S. Neubauer, Sang Eon Park, Melissa Quinnan, Roger Rusack, Zhizhen Zhao

The findable, accessible, interoperable, and reusable (FAIR) data principles have provided a framework for examining, evaluating, and improving how we share data with the aim of facilitating scientific discovery. Efforts have been made to generalize these principles to research software and other digital products. Artificial intelligence (AI) models -- algorithms that have been trained on data rather than explicitly programmed -- are an important target for this because of the ever-increasing pace with which AI is transforming scientific and engineering domains. In this paper, we propose a practical definition of FAIR principles for AI models and create a FAIR AI project template that promotes adherence to these principles. We demonstrate how to implement these principles using a concrete example from experimental high energy physics: a graph neural network for identifying Higgs bosons decaying to bottom quarks. We study the robustness of these FAIR AI models and their portability across hardware architectures and software frameworks, and report new insights on the interpretability of AI predictions by studying the interplay between FAIR datasets and AI models. Enabled by publishing FAIR AI models, these studies pave the way toward reliable and automated AI-driven scientific discovery.

comment: 32 pages, 8 figures, 9 tables

♻ ☆ FAIR principles for AI models with a practical application for accelerated high energy diffraction microscopy

Nikil Ravi, Pranshu Chaturvedi, E. A. Huerta, Zhengchun Liu, Ryan Chard, Aristana Scourtas, K. J. Schmidt, Kyle Chard, Ben Blaiszik, Ian Foster

A concise and measurable set of FAIR (Findable, Accessible, Interoperable and Reusable) principles for scientific data is transforming the state-of-practice for data management and stewardship, supporting and enabling discovery and innovation. Learning from this initiative, and acknowledging the impact of artificial intelligence (AI) in the practice of science and engineering, we introduce a set of practical, concise, and measurable FAIR principles for AI models. We showcase how to create and share FAIR data and AI models within a unified computational framework combining the following elements: the Advanced Photon Source at Argonne National Laboratory, the Materials Data Facility, the Data and Learning Hub for Science, and funcX, and the Argonne Leadership Computing Facility (ALCF), in particular the ThetaGPU supercomputer and the SambaNova DataScale system at the ALCF AI Testbed. We describe how this domain-agnostic computational framework may be harnessed to enable autonomous AI-driven discovery.

comment: 11 pages, 3 figures; Accepted to Scientific Data; for press release see https://www.anl.gov/article/argonne-scientists-promote-fair-standards-for-managing-artificial-intelligence-models and https://www.ncsa.illinois.edu/ncsa-student-researchers-lead-authors-on-award-winning-paper; Received 2022 HPCwire Readers' Choice Award on Best Use of High Performance Data Analytics & Artificial Intelligence

♻ ☆ Berlin V2X: A Machine Learning Dataset from Multiple Vehicles and Radio Access Technologies

Rodrigo Hernangómez, Philipp Geuer, Alexandros Palaios, Daniel Schäufele, Cara Watermann, Khawla Taleb-Bouhemadi, Mohammad Parvini, Anton Krause, Sanket Partani, Christian Vielhaus, Martin Kasparick, Daniel F. Külzer, Friedrich Burmeister, Sławomir Stańczak, Gerhard Fettweis, Hans D. Schotten, Frank H. P. Fitzek

The evolution of wireless communications into 6G and beyond is expected to rely on new machine learning (ML)-based capabilities. These can enable proactive decisions and actions from wireless-network components to sustain quality-of-service (QoS) and user experience. Moreover, new use cases in the area of vehicular and industrial communications will emerge. Specifically in the area of vehicle communication, vehicle-to-everything (V2X) schemes will benefit strongly from such advances. With this in mind, we have conducted a detailed measurement campaign with the purpose of enabling a plethora of diverse ML-based studies. The resulting datasets offer GPS-located wireless measurements across diverse urban environments for both cellular (with two different operators) and sidelink radio access technologies, thus enabling a variety of different studies towards V2X. The datasets are labeled and sampled with a high time resolution. Furthermore, we make the data publicly available with all the necessary information to support the on-boarding of new researchers. We provide an initial analysis of the data showing some of the challenges that ML needs to overcome and the features that ML can leverage, as well as some hints at potential research studies.

comment: 5 pages, 6 figures. Submitted to a conference. Dataset available at https://ieee-dataport.org/open-access/berlin-v2x

♻ ☆ On Characterizing the Trade-off in Invariant Representation Learning

Bashir Sadeghi, Sepehr Dehdashtian, Vishnu Boddeti

Many applications of representation learning, such as privacy preservation, algorithmic fairness, and domain adaptation, desire explicit control over semantic information being discarded. This goal is formulated as satisfying two objectives: maximizing utility for predicting a target attribute while simultaneously being invariant (independent) to a known semantic attribute. Solutions to invariant representation learning (IRepL) problems lead to a trade-off between utility and invariance when they are competing. While existing works study bounds on this trade-off, two questions remain outstanding: 1) What is the exact trade-off between utility and invariance? and 2) What are the encoders (mapping the data to a representation) that achieve the trade-off, and how can we estimate it from training data? This paper addresses these questions for IRepLs in reproducing kernel Hilbert spaces (RKHS)s. Under the assumption that the distribution of a low-dimensional projection of high-dimensional data is approximately normal, we derive a closed-form solution for the global optima of the underlying optimization problem for encoders in RKHSs. This yields closed formulae for a near-optimal trade-off, corresponding optimal representation dimensionality, and the corresponding encoder(s). We also numerically quantify the trade-off on representative problems and compare them to those achieved by baseline IRepL algorithms.

♻ ☆ Does human speech follow Benford's Law?

Leo Hsu, Visar Berisha

Researchers have observed that the frequencies of leading digits in many man-made and naturally occurring datasets follow a logarithmic curve, with digits that start with the number 1 accounting for $\sim 30\%$ of all numbers in the dataset and digits that start with the number 9 accounting for $\sim 5\%$ of all numbers in the dataset. This phenomenon, known as Benford's Law, is highly repeatable and appears in lists of numbers from electricity bills, stock prices, tax returns, house prices, death rates, lengths of rivers, and naturally occurring images. In this paper we demonstrate that human speech spectra also follow Benford's Law on average. That is, when averaged over many speakers, the frequencies of leading digits in speech magnitude spectra follow this distribution, although with some variability at the individual sample level. We use this observation to motivate a new set of features that can be efficiently extracted from speech and demonstrate that these features can be used to classify between human speech and synthetic speech.

♻ ☆ Rank4Class: A Ranking Formulation for Multiclass Classification

Nan Wang, Zhen Qin, Le Yan, Honglei Zhuang, Xuanhui Wang, Michael Bendersky, Marc Najork

Multiclass classification (MCC) is a fundamental machine learning problem of classifying each instance into one of a predefined set of classes. In the deep learning era, extensive efforts have been spent on developing more powerful neural embedding models to better represent the instance for improving MCC performance. In this paper, we do not aim to propose new neural models for instance representation learning, but to show that it is promising to boost MCC performance with a novel formulation through the lens of ranking. In particular, by viewing MCC as to rank classes for an instance, we first argue that ranking metrics, such as Normalized Discounted Cumulative Gain, can be more informative than the commonly used Top-$K$ metrics. We further demonstrate that the dominant neural MCC recipe can be transformed to a neural ranking framework. Based on such generalization, we show that it is intuitive to leverage advanced techniques from the learning to rank literature to improve the MCC performance out of the box. Extensive empirical results on both text and image classification tasks with diverse datasets and backbone neural models show the value of our proposed framework.

♻ ☆ Provably Reliable Large-Scale Sampling from Gaussian Processes NeurIPS 2022

Anthony Stephenson, Robert Allison, Edward Pyzer-Knapp

When comparing approximate Gaussian process (GP) models, it can be helpful to be able to generate data from any GP. If we are interested in how approximate methods perform at scale, we may wish to generate very large synthetic datasets to evaluate them. Na\"{i}vely doing so would cost \(\mathcal{O}(n^3)\) flops and \(\mathcal{O}(n^2)\) memory to generate a size \(n\) sample. We demonstrate how to scale such data generation to large \(n\) whilst still providing guarantees that, with high probability, the sample is indistinguishable from a sample from the desired GP.

comment: Main article 4 pages + 14 pages of supplementary material. To be published in NeurIPS 2022 Proceedings Workshop on "Gaussian Processes, Spatiotemporal Modeling, and Decision-making Systems"

♻ ☆ Efficient Relation-aware Neighborhood Aggregation in Graph Neural Networks via Tensor Decomposition

Peyman Baghershahi, Reshad Hosseini, Hadi Moradi

Many Graph Neural Networks (GNNs) are proposed for KG embedding. However, lots of these methods neglect the importance of the information of relations and combine it with the information of entities inefficiently and mostly additively, leading to low expressiveness. To address this issue, we introduce a general knowledge graph encoder incorporating tensor decomposition in the aggregation function of Relational Graph Convolutional Network (R-GCN). In our model, the parameters of a low-rank core projection tensor, used to transform neighbor entities, are shared across relations to benefit from multi-task learning and produce expressive relation-aware representations. Besides, we propose a low-rank estimation of the core tensor using CP decomposition to compress the model, which is also applicable, as a regularization method, to other similar GNNs. We train our model using a contrastive loss, which relieves the training limitation of the 1-N method on huge graphs. We achieved favorably competitive results on FB15-237 and WN18RR with embeddings in comparably lower dimensions; particularly, we improved R-GCN performance on FB15-237 by 36% with the same decoder.

comment: 11 pages, 4 Tables, 2 figures

♻ ☆ A Novel Plug-and-Play Approach for Adversarially Robust Generalization

Deepak Maurya, Adarsh Barik, Jean Honorio

In this work, we propose a robust framework that employs adversarially robust training to safeguard the machine learning models against perturbed testing data. We achieve this by incorporating the worst-case additive adversarial error within a fixed budget for each sample during model estimation. Our main focus is to provide a plug-and-play solution that can be incorporated in the existing machine learning algorithms with minimal changes. To that end, we derive the ready-to-use solution for several widely used loss functions with a variety of norm constraints on adversarial perturbation for various supervised and unsupervised ML problems, including regression, classification, two-layer neural networks, graphical models, and matrix completion. The solutions are either in closed-form, 1-D optimization, semidefinite programming, difference of convex programming or a sorting-based algorithm. Finally, we validate our approach by showing significant performance improvement on real-world datasets for supervised problems such as regression and classification, as well as for unsupervised problems such as matrix completion and learning graphical models, with very little computational overhead.

♻ ☆ Deconstructing Self-Supervised Monocular Reconstruction: The Design Decisions that Matter

Jaime Spencer, Chris Russell, Simon Hadfield, Richard Bowden

This paper presents an open and comprehensive framework to systematically evaluate state-of-the-art contributions to self-supervised monocular depth estimation. This includes pretraining, backbone, architectural design choices and loss functions. Many papers in this field claim novelty in either architecture design or loss formulation. However, simply updating the backbone of historical systems results in relative improvements of 25%, allowing them to outperform the majority of existing systems. A systematic evaluation of papers in this field was not straightforward. The need to compare like-with-like in previous papers means that longstanding errors in the evaluation protocol are ubiquitous in the field. It is likely that many papers were not only optimized for particular datasets, but also for errors in the data and evaluation criteria. To aid future research in this area, we release a modular codebase (https://github.com/jspenmar/monodepth_benchmark), allowing for easy evaluation of alternate design decisions against corrected data and evaluation criteria. We re-implement, validate and re-evaluate 16 state-of-the-art contributions and introduce a new dataset (SYNS-Patches) containing dense outdoor depth maps in a variety of both natural and urban scenes. This allows for the computation of informative metrics in complex regions such as depth boundaries.

comment: https://github.com/jspenmar/monodepth_benchmark

♻ ☆ Dynamic Budget Throttling in Repeated Second-Price Auctions

Zhaohua Chen, Chang Wang, Qian Wang, Yuqi Pan, Zhuming Shi, Zheng Cai, Yukun Ren, Zhihua Zhu, Xiaotie Deng

In today's online advertising markets, an important demand for an advertiser (buyer) is to control her total expenditure within a time span under some budget. Among all budget control approaches, throttling stands out as a popular one, where the buyer chooses to participate in only a part of auctions. This paper gives a theoretical panorama of a single buyer's dynamic budget throttling process in repeated second-price auctions, which is lacking in the literature. We first establish a lower bound on the regret and an upper bound on the asymptotic competitive ratio for any algorithm, respectively, on whether the buyer's values are stochastic or adversarial. Second, on the algorithmic side, we consider two different information structures, with increasing difficulty in learning the stochastic distribution of the highest competing bid. We further propose the OGD-CB algorithm, which is oblivious to whether the values are stochastic or adversarial and has asymptotically equal results under these two information structures. Specifically, with stochastic values, we demonstrate that this algorithm guarantees a near-optimal expected regret. When values are adversarial, we prove that the proposed algorithm reaches the upper bound on the asymptotic competitive ratio. At last, we compare throttling with pacing, another widely adopted budget control method, in the dynamic setting. In the stochastic case, we illustrate that dynamic pacing is generally better than dynamic throttling for the buyer, which is an extension of known results that dynamic pacing is asymptotically optimal in this scenario. However, in the adversarial case, we give an exciting result indicating that dynamic throttling is the asymptotically optimal dynamic bidding strategy. Our results fill the gaps in the theoretical research of dynamic throttling and comprehensively reveal the ability of this popular budget-smoothing strategy.

comment: 44 pages, 1 figure, 1 table

♻ ☆ Why Deep Learning Generalizes

Benjamin L. Badger

Very large deep learning models trained using gradient descent are remarkably resistant to memorization given their huge capacity, but are at the same time capable of fitting large datasets of pure noise. Here methods are introduced by which models may be trained to memorize datasets that normally are generalized. We find that memorization is difficult relative to generalization, but that adding noise makes memorization easier. Increasing the dataset size exaggerates the characteristics of that dataset: model access to more training samples makes overfitting easier for random data, but somewhat harder for natural images. The bias of deep learning towards generalization is explored theoretically, and we show that generalization results from a model's parameters being attracted to points of maximal stability with respect to that model's inputs during gradient descent.

comment: 11 pages

♻ ☆ Diversity-Promoting Ensemble for Medical Image Segmentation

Mariana-Iuliana Georgescu, Radu Tudor Ionescu, Andreea-Iuliana Miron

Medical image segmentation is an actively studied task in medical imaging, where the precision of the annotations is of utter importance towards accurate diagnosis and treatment. In recent years, the task has been approached with various deep learning systems, among the most popular models being U-Net. In this work, we propose a novel strategy to generate ensembles of different architectures for medical image segmentation, by leveraging the diversity (decorrelation) of the models forming the ensemble. More specifically, we utilize the Dice score among model pairs to estimate the correlation between the outputs of the two models forming each pair. To promote diversity, we select models with low Dice scores among each other. We carry out gastro-intestinal tract image segmentation experiments to compare our diversity-promoting ensemble (DiPE) with another strategy to create ensembles based on selecting the top scoring U-Net models. Our empirical results show that DiPE surpasses both individual models as well as the ensemble creation strategy based on selecting the top scoring models.

comment: Accepted at SAC 2023

♻ ☆ Diffusion Models for Graphs Benefit From Discrete State Spaces NeurIPS 2022

Kilian Konstantin Haefeli, Karolis Martinkus, Nathanaël Perraudin, Roger Wattenhofer

Denoising diffusion probabilistic models and score matching models have proven to be very powerful for generative tasks. While these approaches have also been applied to the generation of discrete graphs, they have, so far, relied on continuous Gaussian perturbations. Instead, in this work, we suggest using discrete noise for the forward Markov process. This ensures that in every intermediate step the graph remains discrete. Compared to the previous approach, our experimental results on four datasets and multiple architectures show that using a discrete noising process results in higher quality generated samples indicated with an average MMDs reduced by a factor of 1.5. Furthermore, the number of denoising steps is reduced from 1000 to 32 steps leading to a 30 times faster sampling procedure.

comment: Presented at the First Learning on Graphs Conference (LoG 2022) and the NeurIPS 2022 New Frontiers in Graph Learning Workshop (NeurIPS GLFrontiers 2022)

♻ ☆ Relative Importance Sampling For Off-Policy Actor-Critic in Deep Reinforcement Learning

Mahammad Humayoo, Xueqi Cheng

Off-policy learning is more unstable compared to on-policy learning in reinforcement learning (RL). One reason for the instability of off-policy learning is a discrepancy between the target ($\pi$) and behavior (b) policy distributions. The discrepancy between $\pi$ and b distributions can be alleviated by employing a smooth variant of the importance sampling (IS), such as the relative importance sampling (RIS). RIS has parameter $\beta\in[0, 1]$ which controls smoothness. To cope with instability, we present the first relative importance sampling-off-policy actor-critic (RIS-Off-PAC) model-free algorithms in RL. In our method, the network yields a target policy (the actor), a value function (the critic) assessing the current policy ($\pi$) using samples drawn from behavior policy. We use action value generated from the behavior policy in reward function to train our algorithm rather than from the target policy. We also use deep neural networks to train both actor and critic. We evaluated our algorithm on a number of Open AI Gym benchmark problems and demonstrate better or comparable performance to several state-of-the-art RL baselines.

♻ ☆ Classification and mapping of low-statured 'shrubland' cover types in post-agricultural landscapes of the US Northeast

Michael J Mahoney, Lucas K Johnson, Abigail Z Guinan, Colin M Beier

Novel plant communities reshape landscapes and pose challenges for land cover classification and mapping that can constrain research and stewardship efforts. In the US Northeast, emergence of low-statured woody vegetation, or shrublands, instead of secondary forests in post-agricultural landscapes is well-documented by field studies, but poorly understood from a landscape perspective, which limits the ability to systematically study and manage these lands. To address gaps in classification/mapping of low-statured cover types where they have been historically rare, we developed models to predict shrubland distributions at 30m resolution across New York State (NYS), using a stacked ensemble combining a random forest, gradient boosting machine, and artificial neural network to integrate remote sensing of structural (airborne LIDAR) and optical (satellite imagery) properties of vegetation cover. We first classified a 1m canopy height model (CHM), derived from a patchwork of available LIDAR coverages, to define shrubland presence/absence. Next, these non-contiguous maps were used to train a model ensemble based on temporally-segmented imagery to predict shrubland probability for the entire study landscape (NYS). Approximately 2.5% of the CHM coverage area was classified as shrubland. Models using Landsat predictors trained on the classified CHM were effective at identifying shrubland (test set AUC=0.893, real-world AUC=0.904), in discriminating between shrub/young forest and other cover classes, and produced qualitatively sensible maps, even when extending beyond the original training data. Our results suggest that incorporation of airborne LiDAR, even from a discontinuous patchwork of coverages, can improve land cover classification of historically rare but increasingly prevalent shrubland habitats across broader areas.

comment: 43 pages (35 main text, 8 supplementary materials); 11 figures (10 main text, 1 supplementary materials), 10 tables (4 main text, 6 supplementary materials)

♻ ☆ Exponentially Improving the Complexity of Simulating the Weisfeiler-Lehman Test with Graph Neural Networks NeurIPS 2022

Anders Aamand, Justin Y. Chen, Piotr Indyk, Shyam Narayanan, Ronitt Rubinfeld, Nicholas Schiefer, Sandeep Silwal, Tal Wagner

Recent work shows that the expressive power of Graph Neural Networks (GNNs) in distinguishing non-isomorphic graphs is exactly the same as that of the Weisfeiler-Lehman (WL) graph test. In particular, they show that the WL test can be simulated by GNNs. However, those simulations involve neural networks for the 'combine' function of size polynomial or even exponential in the number of graph nodes $n$, as well as feature vectors of length linear in $n$. We present an improved simulation of the WL test on GNNs with \emph{exponentially} lower complexity. In particular, the neural network implementing the combine function in each node has only a polylogarithmic number of parameters in $n$, and the feature vectors exchanged by the nodes of GNN consists of only $O(\log n)$ bits. We also give logarithmic lower bounds for the feature vector length and the size of the neural networks, showing the (near)-optimality of our construction.

comment: 22 pages,5 figures, published at NeurIPS 2022. Updated funding statements

♻ ☆ Deep Riemannian Networks for EEG Decoding

Daniel Wilson, Robin Tibor Schirrmeister, Lukas Alexander Wilhelm Gemein, Tonio Ball

State-of-the-art performance in electroencephalography (EEG) decoding tasks is currently often achieved with either Deep-Learning or Riemannian-Geometry-based decoders. Recently, there is growing interest in Deep Riemannian Networks (DRNs) possibly combining the advantages of both previous classes of methods. However, there are still a range of topics where additional insight is needed to pave the way for a more widespread application of DRNs in EEG. These include architecture design questions such as network size and end-to-end ability as well as model training questions. How these factors affect model performance has not been explored. Additionally, it is not clear how the data within these networks is transformed, and whether this would correlate with traditional EEG decoding. Our study aims to lay the groundwork in the area of these topics through the analysis of DRNs for EEG with a wide range of hyperparameters. Networks were tested on two public EEG datasets and compared with state-of-the-art ConvNets. Here we propose end-to-end EEG SPDNet (EE(G)-SPDNet), and we show that this wide, end-to-end DRN can outperform the ConvNets, and in doing so use physiologically plausible frequency regions. We also show that the end-to-end approach learns more complex filters than traditional band-pass filters targeting the classical alpha, beta, and gamma frequency bands of the EEG, and that performance can benefit from channel specific filtering approaches. Additionally, architectural analysis revealed areas for further improvement due to the possible loss of Riemannian specific information throughout the network. Our study thus shows how to design and train DRNs to infer task-related information from the raw EEG without the need of handcrafted filterbanks and highlights the potential of end-to-end DRNs such as EE(G)-SPDNet for high-performance EEG decoding.

comment: 26 pages, 15 Figures

♻ ☆ Towards biologically plausible Dreaming and Planning in recurrent spiking networks

Cristiano Capone, Pier Stanislao Paolucci

Humans and animals can learn new skills after practicing for a few hours, while current reinforcement learning algorithms require a large amount of data to achieve good performances. Recent model-based approaches show promising results by reducing the number of necessary interactions with the environment to learn a desirable policy. However, these methods require biological implausible ingredients, such as the detailed storage of older experiences, and long periods of offline learning. The optimal way to learn and exploit word-models is still an open question. Taking inspiration from biology, we suggest that dreaming might be an efficient expedient to use an inner model. We propose a two-module (agent and model) spiking neural network in which "dreaming" (living new experiences in a model-based simulated environment) significantly boosts learning. We also explore "planning", an online alternative to dreaming, that shows comparable performances. Importantly, our model does not require the detailed storage of experiences, and learns online the world-model and the policy. Moreover, we stress that our network is composed of spiking neurons, further increasing the biological plausibility and implementability in neuromorphic hardware.

♻ ☆ MutexMatch: Semi-Supervised Learning with Mutex-Based Consistency Regularization

Yue Duan, Zhen Zhao, Lei Qi, Lei Wang, Luping Zhou, Yinghuan Shi, Yang Gao

The core issue in semi-supervised learning (SSL) lies in how to effectively leverage unlabeled data, whereas most existing methods tend to put a great emphasis on the utilization of high-confidence samples yet seldom fully explore the usage of low-confidence samples. In this paper, we aim to utilize low-confidence samples in a novel way with our proposed mutex-based consistency regularization, namely MutexMatch. Specifically, the high-confidence samples are required to exactly predict "what it is" by conventional True-Positive Classifier, while the low-confidence samples are employed to achieve a simpler goal -- to predict with ease "what it is not" by True-Negative Classifier. In this sense, we not only mitigate the pseudo-labeling errors but also make full use of the low-confidence unlabeled data by consistency of dissimilarity degree. MutexMatch achieves superior performance on multiple benchmark datasets, i.e., CIFAR-10, CIFAR-100, SVHN, STL-10, mini-ImageNet and Tiny-ImageNet. More importantly, our method further shows superiority when the amount of labeled data is scarce, e.g., 92.23% accuracy with only 20 labeled data on CIFAR-10. Our code and model weights have been released at https://github.com/NJUyued/MutexMatch4SSL.

comment: Accepted by IEEE Transactions on Neural Networks and Learning Systems (TNNLS)

♻ ☆ Towards Linguistically Informed Multi-Objective Pre-Training for Natural Language Inference

Maren Pielka, Svetlana Schmidt, Lisa Pucknat, Rafet Sifa

We introduce a linguistically enhanced combination of pre-training methods for transformers. The pre-training objectives include POS-tagging, synset prediction based on semantic knowledge graphs, and parent prediction based on dependency parse trees. Our approach achieves competitive results on the Natural Language Inference task, compared to the state of the art. Specifically for smaller models, the method results in a significant performance boost, emphasizing the fact that intelligent pre-training can make up for fewer parameters and help building more efficient models. Combining POS-tagging and synset prediction yields the overall best results.

♻ ☆ ADAS: A Simple Active-and-Adaptive Baseline for Cross-Domain 3D Semantic Segmentation

Ben Fei, Siyuan Huang, Jiakang Yuan, Botian Shi, Bo Zhang, Tao Chen, Min Dou, Yu Qiao

State-of-the-art 3D semantic segmentation models are trained on the off-the-shelf public benchmarks, but they often face the major challenge when these well-trained models are deployed to a new domain. In this paper, we propose an Active-and-Adaptive Segmentation (ADAS) baseline to enhance the weak cross-domain generalization ability of a well-trained 3D segmentation model, and bridge the point distribution gap between domains. Specifically, before the cross-domain adaptation stage begins, ADAS performs an active sampling operation to select a maximally-informative subset from both source and target domains for effective adaptation, reducing the adaptation difficulty under 3D scenarios. Benefiting from the rise of multi-modal 2D-3D datasets, ADAS utilizes a cross-modal attention-based feature fusion module that can extract a representative pair of image features and point features to achieve a bi-directional image-point feature interaction for better safe adaptation. Experimentally, ADAS is verified to be effective in many cross-domain settings including: 1) Unsupervised Domain Adaptation (UDA), which means that all samples from target domain are unlabeled; 2) Unsupervised Few-shot Domain Adaptation (UFDA) which means that only a few unlabeled samples are available in the unlabeled target domain; 3) Active Domain Adaptation (ADA) which means that the selected target samples by ADAS are manually annotated. Their results demonstrate that ADAS achieves a significant accuracy gain by easily coupling ADAS with self-training methods or off-the-shelf UDA works.

♻ ☆ Reconstruction of Multivariate Sparse Signals from Mismatched Samples

Taulant Koka, Michael Muma, Benjamín Béjar Haro

Erroneous correspondences between samples and their respective channel or target commonly arise in several real-world applications. For instance, whole-brain calcium imaging of freely moving organisms, multiple target tracking or multi-person contactless vital sign monitoring may be severely affected by mismatched sample-channel assignments. To systematically address this fundamental problem, we pose it as a signal reconstruction problem where we have lost correspondences between the samples and their respective channels. We show that under the assumption that the signals of interest admit a sparse representation over an overcomplete dictionary, unique signal recovery is possible. Our derivations reveal that the problem is equivalent to a structured unlabeled sensing problem without precise knowledge of the sensing matrix. Unfortunately, existing methods are neither robust to errors in the regressors nor do they exploit the structure of the problem. Therefore, we propose a novel robust two-step approach for the reconstruction of shuffled sparse signals. The performance and robustness of the proposed approach is illustrated in an application of whole-brain calcium imaging in computational neuroscience. The proposed framework can be generalized to sparse signal representations other than the ones considered in this work to be applied in a variety of real-world problems with imprecise measurement or channel assignment.

comment: There is an error in the use of Corollary 1 in our Paper, which does not apply in our case

♻ ☆ Learning Spectral Unions of Partial Deformable 3D Shapes

Luca Moschella, Simone Melzi, Luca Cosmo, Filippo Maggioli, Or Litany, Maks Ovsjanikov, Leonidas Guibas, Emanuele Rodolà

Spectral geometric methods have brought revolutionary changes to the field of geometry processing. Of particular interest is the study of the Laplacian spectrum as a compact, isometry and permutation-invariant representation of a shape. Some recent works show how the intrinsic geometry of a full shape can be recovered from its spectrum, but there are approaches that consider the more challenging problem of recovering the geometry from the spectral information of partial shapes. In this paper, we propose a possible way to fill this gap. We introduce a learning-based method to estimate the Laplacian spectrum of the union of partial non-rigid 3D shapes, without actually computing the 3D geometry of the union or any correspondence between those partial shapes. We do so by operating purely in the spectral domain and by defining the union operation between short sequences of eigenvalues. We show that the approximated union spectrum can be used as-is to reconstruct the complete geometry [MRC*19], perform region localization on a template [RTO*19] and retrieve shapes from a database, generalizing ShapeDNA [RWP06] to work with partialities. Working with eigenvalues allows us to deal with unknown correspondence, different sampling, and different discretizations (point clouds and meshes alike), making this operation especially robust and general. Our approach is data-driven and can generalize to isometric and non-isometric deformations of the surface, as long as these stay within the same semantic class (e.g., human bodies or horses), as well as to partiality artifacts not seen at training time.

comment: 18 pages, 20 figures

♻ ☆ Expander Graph Propagation NeurIPS 2022

Andreea Deac, Marc Lackenby, Petar Veličković

Deploying graph neural networks (GNNs) on whole-graph classification or regression tasks is known to be challenging: it often requires computing node features that are mindful of both local interactions in their neighbourhood and the global context of the graph structure. GNN architectures that navigate this space need to avoid pathological behaviours, such as bottlenecks and oversquashing, while ideally having linear time and space complexity requirements. In this work, we propose an elegant approach based on propagating information over expander graphs. We leverage an efficient method for constructing expander graphs of a given size, and use this insight to propose the EGP model. We show that EGP is able to address all of the above concerns, while requiring minimal effort to set up, and provide evidence of its empirical utility on relevant graph classification datasets and baselines in the Open Graph Benchmark. Importantly, using expander graphs as a template for message passing necessarily gives rise to negative curvature. While this appears to be counterintuitive in light of recent related work on oversquashing, we theoretically demonstrate that negatively curved edges are likely to be required to obtain scalable message passing without bottlenecks. To the best of our knowledge, this is a previously unstudied result in the context of graph representation learning, and we believe our analysis paves the way to a novel class of scalable methods to counter oversquashing in GNNs.

comment: Presented at LoG 2022. Best Paper Award at the NeurIPS 2022 Workshop on New Frontiers in Graph Learning (GLFrontiers). 18 pages, 1 figure

♻ ☆ Time to augment self-supervised visual representation learning

Arthur Aubret, Markus Ernst, Céline Teulière, Jochen Triesch

Biological vision systems are unparalleled in their ability to learn visual representations without supervision. In machine learning, self-supervised learning (SSL) has led to major advances in forming object representations in an unsupervised fashion. Such systems learn representations invariant to augmentation operations over images, like cropping or flipping. In contrast, biological vision systems exploit the temporal structure of the visual experience during natural interactions with objects. This gives access to "augmentations" not commonly used in SSL, like watching the same object from multiple viewpoints or against different backgrounds. Here, we systematically investigate and compare the potential benefits of such time-based augmentations during natural interactions for learning object categories. Our results show that time-based augmentations achieve large performance gains over state-of-the-art image augmentations. Specifically, our analyses reveal that: 1) 3-D object manipulations drastically improve the learning of object categories; 2) viewing objects against changing backgrounds is important for learning to discard background-related information from the latent representation. Overall, we conclude that time-based augmentations during natural interactions with objects can substantially improve self-supervised learning, narrowing the gap between artificial and biological vision systems.

comment: 20 pages

♻ ☆ Federated Graph Neural Networks: Overview, Techniques and Challenges

Rui Liu, Pengwei Xing, Zichao Deng, Anran Li, Cuntai Guan, Han Yu

With its capability to deal with graph data, which is widely found in practical applications, graph neural networks (GNNs) have attracted significant research attention in recent years. As societies become increasingly concerned with the need for data privacy protection, GNNs face the need to adapt to this new normal. Besides, as clients in Federated Learning (FL) may have relationships, more powerful tools are required to utilize such implicit information to boost performance. This has led to the rapid development of the emerging research field of federated graph neural networks (FedGNNs). This promising interdisciplinary field is highly challenging for interested researchers to grasp. The lack of an insightful survey on this topic further exacerbates the entry difficulty. In this paper, we bridge this gap by offering a comprehensive survey of this emerging field. We propose a 2-dimensional taxonomy of the FedGNNs literature: 1) the main taxonomy provides a clear perspective on the integration of GNNs and FL by analyzing how GNNs enhance FL training as well as how FL assists GNNs training, and 2) the auxiliary taxonomy provides a view on how FedGNNs deal with heterogeneity across FL clients. Through discussions of key ideas, challenges, and limitations of existing works, we envision future research directions that can help build more robust, explainable, efficient, fair, inductive, and comprehensive FedGNNs.

♻ ☆ Order Optimal Bounds for One-Shot Federated Learning over non-Convex Loss Functions

Arsalan Sharifnassab, Saber Salehkaleybar, S. Jamaloddin Golestani

We consider the problem of federated learning in a one-shot setting in which there are $m$ machines, each observing $n$ sample functions from an unknown distribution on non-convex loss functions. Let $F:[-1,1]^d\rightarrow\mathbb{R}$ be the expected loss function with respect to this unknown distribution. The goal is to find an estimate of the minimizer of $F$. Based on its observations, each machine generates a signal of bounded length $B$ and sends it to a server. The server collects signals of all machines and outputs an estimate of the minimizer of $F$. We show that the expected loss of any algorithm is lower bounded by $\max\big(1/(\sqrt{n}(mB)^{1/d}), 1/\sqrt{mn}\big)$, up to a logarithmic factor. We then prove that this lower bound is order optimal in $m$ and $n$ by presenting a distributed learning algorithm, called Multi-Resolution Estimator for Non-Convex loss function (MRE-NC), whose expected loss matches the lower bound for large $mn$ up to polylogarithmic factors.

♻ ☆ EZNAS: Evolving Zero Cost Proxies For Neural Architecture Scoring

Yash Akhauri, J. Pablo Munoz, Nilesh Jain, Ravi Iyer

Neural Architecture Search (NAS) has significantly improved productivity in the design and deployment of neural networks (NN). As NAS typically evaluates multiple models by training them partially or completely, the improved productivity comes at the cost of significant carbon footprint. To alleviate this expensive training routine, zero-shot/cost proxies analyze an NN at initialization to generate a score, which correlates highly with its true accuracy. Zero-cost proxies are currently designed by experts conducting multiple cycles of empirical testing on possible algorithms, datasets, and neural architecture design spaces. This experimentation lowers productivity and is an unsustainable approach towards zero-cost proxy design as deep learning use-cases diversify in nature. Additionally, existing zero-cost proxies fail to generalize across neural architecture design spaces. In this paper, we propose a genetic programming framework to automate the discovery of zero-cost proxies for neural architecture scoring. Our methodology efficiently discovers an interpretable and generalizable zero-cost proxy that gives state of the art score-accuracy correlation on all datasets and search spaces of NASBench-201 and Network Design Spaces (NDS). We believe that this research indicates a promising direction towards automatically discovering zero-cost proxies that can work across network architecture design spaces, datasets, and tasks.

♻ ☆ ColoristaNet for Photorealistic Video Style Transfer

Xiaowen Qiu, Ruize Xu, Boan He, Yingtao Zhang, Wenqiang Zhang, Weifeng Ge

Photorealistic style transfer aims to transfer the artistic style of an image onto an input image or video while keeping photorealism. In this paper, we think it's the summary statistics matching scheme in existing algorithms that leads to unrealistic stylization. To avoid employing the popular Gram loss, we propose a self-supervised style transfer framework, which contains a style removal part and a style restoration part. The style removal network removes the original image styles, and the style restoration network recovers image styles in a supervised manner. Meanwhile, to address the problems in current feature transformation methods, we propose decoupled instance normalization to decompose feature transformation into style whitening and restylization. It works quite well in ColoristaNet and can transfer image styles efficiently while keeping photorealism. To ensure temporal coherency, we also incorporate optical flow methods and ConvLSTM to embed contextual information. Experiments demonstrates that ColoristaNet can achieve better stylization effects when compared with state-of-the-art algorithms.

comment: 30 pages, 29 figures

♻ ☆ MMDialog: A Large-scale Multi-turn Dialogue Dataset Towards Multi-modal Open-domain Conversation

Jiazhan Feng, Qingfeng Sun, Can Xu, Pu Zhao, Yaming Yang, Chongyang Tao, Dongyan Zhao, Qingwei Lin

Responding with multi-modal content has been recognized as an essential capability for an intelligent conversational agent. In this paper, we introduce the MMDialog dataset to better facilitate multi-modal conversation. MMDialog is composed of a curated set of 1.08 million real-world dialogues with 1.53 million unique images across 4,184 topics. MMDialog has two main and unique advantages. First, it is the largest multi-modal conversation dataset by the number of dialogues by 88x. Second, it contains massive topics to generalize the open-domain. To build engaging dialogue system with this dataset, we propose and normalize two response producing tasks based on retrieval and generative scenarios. In addition, we build two baselines for above tasks with state-of-the-art techniques and report their experimental performance. We also propose a novel evaluation metric MM-Relevance to measure the multi-modal responses. Our dataset and scripts are available in https://github.com/victorsungo/MMDialog.

♻ ☆ SMSMix: Sense-Maintained Sentence Mixup for Word Sense Disambiguation EMNLP2022

Hee Suk Yoon, Eunseop Yoon, John Harvill, Sunjae Yoon, Mark Hasegawa-Johnson, Chang D. Yoo

Word Sense Disambiguation (WSD) is an NLP task aimed at determining the correct sense of a word in a sentence from discrete sense choices. Although current systems have attained unprecedented performances for such tasks, the nonuniform distribution of word senses during training generally results in systems performing poorly on rare senses. To this end, we consider data augmentation to increase the frequency of these least frequent senses (LFS) to reduce the distributional bias of senses during training. We propose Sense-Maintained Sentence Mixup (SMSMix), a novel word-level mixup method that maintains the sense of a target word. SMSMix smoothly blends two sentences using mask prediction while preserving the relevant span determined by saliency scores to maintain a specific word's sense. To the best of our knowledge, this is the first attempt to apply mixup in NLP while preserving the meaning of a specific word. With extensive experiments, we validate that our augmentation method can effectively give more information about rare senses during training with maintained target sense label.

comment: EMNLP2022

♻ ☆ Answering Complex Logical Queries on Knowledge Graphs via Query Computation Tree Optimization

Yushi Bai, Xin Lv, Juanzi Li, Lei Hou

Answering complex logical queries on incomplete knowledge graphs is a challenging task, and has been widely studied. Embedding-based methods require training on complex queries, and cannot generalize well to out-of-distribution query structures. Recent work frames this task as an end-to-end optimization problem, and it only requires a pretrained link predictor. However, due to the exponentially large combinatorial search space, the optimal solution can only be approximated, limiting the final accuracy. In this work, we propose QTO (Query Computation Tree Optimization) that can efficiently find the exact optimal solution. QTO finds the optimal solution by a forward-backward propagation on the tree-like computation graph, i.e., query computation tree. In particular, QTO utilizes the independence encoded in the query computation tree to reduce the search space, where only local computations are involved during the optimization procedure. Experiments on 3 datasets show that QTO obtains state-of-the-art performance on complex query answering, outperforming previous best results by an average of 22%. Moreover, QTO can interpret the intermediate solutions for each of the one-hop atoms in the query with over 90% accuracy.

comment: Code is available at https://github.com/bys0318/QTO

♻ ☆ Lie Group Forced Variational Integrator Networks for Learning and Control of Robot Systems

Valentin Duruisseaux, Thai Duong, Melvin Leok, Nikolay Atanasov

Incorporating prior knowledge of physics laws and structural properties of dynamical systems into the design of deep learning architectures has proven to be a powerful technique for improving their computational efficiency and generalization capacity. Learning accurate models of robot dynamics is critical for safe and stable control. Autonomous mobile robots, including wheeled, aerial, and underwater vehicles, can be modeled as controlled Lagrangian or Hamiltonian rigid-body systems evolving on matrix Lie groups. In this paper, we introduce a new structure-preserving deep learning architecture, the Lie group Forced Variational Integrator Network (LieFVIN), capable of learning controlled Lagrangian or Hamiltonian dynamics on Lie groups, either from position-velocity or position-only data. By design, LieFVINs preserve both the Lie group structure on which the dynamics evolve and the symplectic structure underlying the Hamiltonian or Lagrangian systems of interest. The proposed architecture learns surrogate discrete-time flow maps allowing accurate and fast prediction without numerical-integrator, neural-ODE, or adjoint techniques, which are needed for vector fields. Furthermore, the learnt discrete-time dynamics can be utilized with computationally scalable discrete-time (optimal) control strategies.

comment: 21 pages

♻ ☆ Multi-branch Cascaded Swin Transformers with Attention to k-space Sampling Pattern for Accelerated MRI Reconstruction

Mevan Ekanayake, Kamlesh Pawar, Mehrtash Harandi, Gary Egan, Zhaolin Chen

Global correlations are widely seen in human anatomical structures due to similarity across tissues and bones. These correlations are reflected in magnetic resonance imaging (MRI) scans as a result of close-range proton density and T1/T2 parameters. Furthermore, to achieve accelerated MRI, k-space data are undersampled which causes global aliasing artifacts. Convolutional neural network (CNN) models are widely utilized for accelerated MRI reconstruction, but those models are limited in capturing global correlations due to the intrinsic locality of the convolution operation. The self-attention-based transformer models are capable of capturing global correlations among image features, however, the current contributions of transformer models for MRI reconstruction are minute. The existing contributions mostly provide CNN-transformer hybrid solutions and rarely leverage the physics of MRI. In this paper, we propose a physics-based stand-alone (convolution free) transformer model titled, the Multi-head Cascaded Swin Transformers (McSTRA) for accelerated MRI reconstruction. McSTRA combines several interconnected MRI physics-related concepts with the transformer networks: it exploits global MR features via the shifted window self-attention mechanism; it extracts MR features belonging to different spectral components separately using a multi-head setup; it iterates between intermediate de-aliasing and k-space correction via a cascaded network with data consistency in k-space and intermediate loss computations; furthermore, we propose a novel positional embedding generation mechanism to guide self-attention utilizing the point spread function corresponding to the undersampling mask. Our model significantly outperforms state-of-the-art MRI reconstruction methods both visually and quantitatively while depicting improved resolution and removal of aliasing artifacts.

♻ ☆ On the Convergence of Momentum-Based Algorithms for Federated Bilevel Optimization Problems

Hongchang Gao

In this paper, we studied the federated bilevel optimization problem, which has widespread applications in machine learning. In particular, we developed two momentum-based algorithms for optimizing this kind of problem and established the convergence rate of our two algorithms, providing the sample and communication complexities. Importantly, to the best of our knowledge, our convergence rate is the first one achieving the linear speedup with respect to the number of devices for federated bilevel optimization algorithms. At last, our extensive experimental results confirm the effectiveness of our two algorithms.

♻ ☆ Two-view Graph Neural Networks for Knowledge Graph Completion

Vinh Tong, Dai Quoc Nguyen, Dinh Phung, Dat Quoc Nguyen

We present an effective graph neural network (GNN)-based knowledge graph embedding model, which we name WGE, to capture entity- and relation-focused graph structures. Given a knowledge graph, WGE builds a single undirected entity-focused graph that views entities as nodes. WGE also constructs another single undirected graph from relation-focused constraints, which views entities and relations as nodes. WGE then proposes a GNN-based architecture to better learn vector representations of entities and relations from these two single entity- and relation-focused graphs. WGE feeds the learned entity and relation representations into a weighted score function to return the triple scores for knowledge graph completion. Experimental results show that WGE outperforms competitive baselines, obtaining state-of-the-art performances on seven benchmark datasets for knowledge graph completion.

comment: 17 pages; 4 tables; 4 figures

♻ ☆ FedDAG: Federated DAG Structure Learning

Erdun Gao, Junjia Chen, Li Shen, Tongliang Liu, Mingming Gong, Howard Bondell

To date, most directed acyclic graphs (DAGs) structure learning approaches require data to be stored in a central server. However, due to the consideration of privacy protection, data owners gradually refuse to share their personalized raw data to avoid private information leakage, making this task more troublesome by cutting off the first step. Thus, a puzzle arises: \textit{how do we discover the underlying DAG structure from decentralized data?} In this paper, focusing on the additive noise models (ANMs) assumption of data generation, we take the first step in developing a gradient-based learning framework named FedDAG, which can learn the DAG structure without directly touching the local data and also can naturally handle the data heterogeneity. Our method benefits from a two-level structure of each local model. The first level structure learns the edges and directions of the graph and communicates with the server to get the model information from other clients during the learning procedure, while the second level structure approximates the mechanisms among variables and personally updates on its own data to accommodate the data heterogeneity. Moreover, FedDAG formulates the overall learning task as a continuous optimization problem by taking advantage of an equality acyclicity constraint, which can be solved by gradient descent methods to boost the searching efficiency. Extensive experiments on both synthetic and real-world datasets verify the efficacy of the proposed method.

comment: Accepted to Transactions on Machine Learning Research

♻ ☆ Unsupervised Learning of Neurosymbolic Encoders

Eric Zhan, Jennifer J. Sun, Ann Kennedy, Yisong Yue, Swarat Chaudhuri

We present a framework for the unsupervised learning of neurosymbolic encoders, which are encoders obtained by composing neural networks with symbolic programs from a domain-specific language. Our framework naturally incorporates symbolic expert knowledge into the learning process, which leads to more interpretable and factorized latent representations compared to fully neural encoders. We integrate modern program synthesis techniques with the variational autoencoding (VAE) framework, in order to learn a neurosymbolic encoder in conjunction with a standard decoder. The programmatic descriptions from our encoders can benefit many analysis workflows, such as in behavior modeling where interpreting agent actions and movements is important. We evaluate our method on learning latent representations for real-world trajectory data from animal biology and sports analytics. We show that our approach offers significantly better separation of meaningful categories than standard VAEs and leads to practical gains on downstream analysis tasks, such as for behavior classification.

♻ ☆ Multi-View Active Learning for Short Text Classification in User-Generated Data EMNLP

Payam Karisani, Negin Karisani, Li Xiong

Mining user-generated data often suffers from the lack of enough labeled data, short document lengths, and the informal user language. In this paper, we propose a novel active learning model to overcome these obstacles in the tasks tailored for query phrases--e.g., detecting positive reports of natural disasters. Our model has three novelties: 1) It is the first approach to employ multi-view active learning in this domain. 2) It uses the Parzen-Rosenblatt window method to integrate the representativeness measure into multi-view active learning. 3) It employs a query-by-committee strategy, based on the agreement between predictors, to address the usually noisy language of the documents in this domain. We evaluate our model in four publicly available Twitter datasets with distinctly different applications. We also compare our model with a wide range of baselines including those with multiple classifiers. The experiments testify that our model is highly consistent and outperforms existing models.

comment: EMNLP Findings 2022

♻ ☆ Synthesizing Informative Training Samples with GAN NeurIPS 2022

Bo Zhao, Hakan Bilen

Remarkable progress has been achieved in synthesizing photo-realistic images with generative adversarial networks (GANs). Recently, GANs are utilized as the training sample generator when obtaining or storing real training data is expensive even infeasible. However, traditional GANs generated images are not as informative as the real training samples when being used to train deep neural networks. In this paper, we propose a novel method to synthesize Informative Training samples with GAN (IT-GAN). Specifically, we freeze a pre-trained GAN model and learn the informative latent vectors that correspond to informative training samples. The synthesized images are required to preserve information for training deep neural networks rather than visual reality or fidelity. Experiments verify that the deep neural networks can learn faster and achieve better performance when being trained with our IT-GAN generated images. We also show that our method is a promising solution to dataset condensation problem.

comment: NeurIPS 2022 Workshop on Synthetic Data for Empowering ML Research, https://openreview.net/forum?id=frAv0jtUMfS

♻ ☆ Named Tensor Notation

David Chiang, Alexander M. Rush, Boaz Barak

We propose a notation for tensors with named axes, which relieves the author, reader, and future implementers of machine learning models from the burden of keeping track of the order of axes and the purpose of each. The notation makes it easy to lift operations on low-order tensors to higher order ones, for example, from images to minibatches of images, or from an attention mechanism to multiple attention heads. After a brief overview and formal definition of the notation, we illustrate it through several examples from modern machine learning, from building blocks like attention and convolution to full models like Transformers and LeNet. We then discuss differential calculus in our notation and compare with some alternative notations. Our proposals build on ideas from many previous papers and software libraries. We hope that our notation will encourage more authors to use named tensors, resulting in clearer papers and more precise implementations.

♻ ☆ Variable Clustering via Distributionally Robust Nodewise Regression

Kaizheng Wang, Xiao Xu, Xun Yu Zhou

We study a multi-factor block model for variable clustering and connect it to the regularized subspace clustering by formulating a distributionally robust version of the nodewise regression. To solve the latter problem, we derive a convex relaxation, provide guidance on selecting the size of the robust region, and hence the regularization weighting parameter, based on the data, and propose an ADMM algorithm for implementation. We validate our method in an extensive simulation study. Finally, we propose and apply a variant of our method to stock return data, obtain interpretable clusters that facilitate portfolio selection and compare its out-of-sample performance with other clustering methods in an empirical study.

comment: 34 pages

♻ ☆ AnchorGAE: General Data Clustering via $O(n)$ Bipartite Graph Convolution

Hongyuan Zhang, Jiankun Shi, Rui Zhang, Xuelong Li

Since the representative capacity of graph-based clustering methods is usually limited by the graph constructed on the original features, it is attractive to find whether graph neural networks (GNNs) can be applied to augment the capacity. The core problems mainly come from two aspects: (1) the graph is unavailable in the most clustering scenes so that how to construct high-quality graphs on the non-graph data is usually the most important part; (2) given n samples, the graph-based clustering methods usually consume at least $\mathcal O(n^2)$ time to build graphs and the graph convolution requires nearly $\mathcal O(n^2)$ for a dense graph and $\mathcal O(|\mathcal{E}|)$ for a sparse one with $|\mathcal{E}|$ edges. Accordingly, both graph-based clustering and GNNs suffer from the severe inefficiency problem. To tackle these problems, we propose a novel clustering method, AnchorGAE, with the self-supervised estimation of graph and efficient graph convolution. We first show how to convert a non-graph dataset into a graph dataset, by introducing the generative graph model and anchors. We then show that the constructed bipartite graph can reduce the computational complexity of graph convolution from $\mathcal O(n^2)$ and $\mathcal O(|\mathcal{E}|)$ to $\mathcal O(n)$. The succeeding steps for clustering can be easily designed as $\mathcal O(n)$ operations. Interestingly, the anchors naturally lead to siamese architecture with the help of the Markov process. Furthermore, the estimated bipartite graph is updated dynamically according to the features extracted by GNN, to promote the quality of the graph. However, we theoretically prove that the self-supervised paradigm frequently results in a collapse that often occurs after 2-3 update iterations in experiments, especially when the model is well-trained. A specific strategy is accordingly designed to prevent the collapse.

comment: copyright 2022 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works

♻ ☆ Functional Linear Regression of Cumulative Distribution Functions

Qian Zhang, Anuran Makur, Kamyar Azizzadenesheli

The estimation of cumulative distribution functions (CDFs) is an important learning task with a great variety of downstream applications, such as risk assessments in predictions and decision making. In this paper, we study functional regression of contextual CDFs where each data point is sampled from a linear combination of context dependent CDF basis functions. We propose functional ridge-regression-based estimation methods that estimate CDFs accurately everywhere. In particular, given $n$ samples with $d$ basis functions, we show estimation error upper bounds of $\widetilde{O}(\sqrt{d/n})$ for fixed design, random design, and adversarial context cases. We also derive matching information theoretic lower bounds, establishing minimax optimality for CDF functional regression. Furthermore, we remove the burn-in time in the random design setting using an alternative penalized estimator. Then, we consider agnostic settings where there is a mismatch in the data generation process. We characterize the error of the proposed estimators in terms of the mismatched error, and show that the estimators are well-behaved under model mismatch. Finally, to complete our study, we formalize infinite dimensional models where the parameter space is an infinite dimensional Hilbert space, and establish self-normalized estimation error upper bounds for this setting.

comment: 62 pages, 2 figures

♻ ☆ Decision-making at Unsignalized Intersection for Autonomous Vehicles: Left-turn Maneuver with Deep Reinforcement Learning

Feng Wang, Dongjie Shi, Teng Liu, Xiaolin Tang

Decision-making module enables autonomous vehicles to reach appropriate maneuvers in the complex urban environments, especially the intersection situations. This work proposes a deep reinforcement learning (DRL) based left-turn decision-making framework at unsignalized intersection for autonomous vehicles. The objective of the studied automated vehicle is to make an efficient and safe left-turn maneuver at a four-way unsignalized intersection. The exploited DRL methods include deep Q-learning (DQL) and double DQL. Simulation results indicate that the presented decision-making strategy could efficaciously reduce the collision rate and improve transport efficiency. This work also reveals that the constructed left-turn control structure has a great potential to be applied in real-time.

comment: 17 pages, 10 figures

♻ ☆ Planning with Diffusion for Flexible Behavior Synthesis ICML 2022

Michael Janner, Yilun Du, Joshua B. Tenenbaum, Sergey Levine

Model-based reinforcement learning methods often use learning only for the purpose of estimating an approximate dynamics model, offloading the rest of the decision-making work to classical trajectory optimizers. While conceptually simple, this combination has a number of empirical shortcomings, suggesting that learned models may not be well-suited to standard trajectory optimization. In this paper, we consider what it would look like to fold as much of the trajectory optimization pipeline as possible into the modeling problem, such that sampling from the model and planning with it become nearly identical. The core of our technical approach lies in a diffusion probabilistic model that plans by iteratively denoising trajectories. We show how classifier-guided sampling and image inpainting can be reinterpreted as coherent planning strategies, explore the unusual and useful properties of diffusion-based planning methods, and demonstrate the effectiveness of our framework in control settings that emphasize long-horizon decision-making and test-time flexibility.

comment: ICML 2022 (long talk). Project page and code at https://diffusion-planning.github.io/

♻ ☆ An Extensive Data Processing Pipeline for MIMIC-IV

Mehak Gupta, Brennan Gallamoza, Nicolas Cutrona, Pranjal Dhakal, Raphael Poulain, Rahmatollah Beheshti

An increasing amount of research is being devoted to applying machine learning methods to electronic health record (EHR) data for various clinical purposes. This growing area of research has exposed the challenges of the accessibility of EHRs. MIMIC is a popular, public, and free EHR dataset in a raw format that has been used in numerous studies. The absence of standardized pre-processing steps can be, however, a significant barrier to the wider adoption of this rare resource. Additionally, this absence can reduce the reproducibility of the developed tools and limit the ability to compare the results among similar studies. In this work, we provide a greatly customizable pipeline to extract, clean, and pre-process the data available in the fourth version of the MIMIC dataset (MIMIC-IV). The pipeline also presents an end-to-end wizard-like package supporting predictive model creations and evaluations. The pipeline covers a range of clinical prediction tasks which can be broadly classified into four categories - readmission, length of stay, mortality, and phenotype prediction. The tool is publicly available at https://github.com/healthylaife/MIMIC-IV-Data-Pipeline.

♻ ☆ Simple Neighborhood Representative Pre-processing Boosts Outlier Detectors

Jiawei Yang, Yu Chen, Sylwan Rahardja

Over the decades, traditional outlier detectors have ignored the group-level factor when calculating outlier scores for objects in data by evaluating only the object-level factor, failing to capture the collective outliers. To mitigate this issue, we present a method called neighborhood representative (NR), which empowers all the existing outlier detectors to efficiently detect outliers, including collective outliers, while maintaining their computational integrity. It achieves this by selecting representative objects, scoring these objects, then applies the score of the representative objects to its collective objects. Without altering existing detectors, NR is compatible with existing detectors, while improving performance on real world datasets with +8% (0.72 to 0.78 AUC) relative to state-of-the-art outlier detectors.

♻ ☆ Sliced Optimal Partial Transport

Yikun Bai, Bernard Schmitzer, Mathew Thorpe, Soheil Kolouri

Optimal transport (OT) has become exceedingly popular in machine learning, data science, and computer vision. The core assumption in the OT problem is the equal total amount of mass in source and target measures, which limits its application. Optimal Partial Transport (OPT) is a recently proposed solution to this limitation. Similar to the OT problem, the computation of OPT relies on solving a linear programming problem (often in high dimensions), which can become computationally prohibitive. In this paper, we propose an efficient algorithm for calculating the OPT problem between two non-negative measures in one dimension. Next, following the idea of sliced OT distances, we utilize slicing to define the sliced OPT distance. Finally, we demonstrate the computational and accuracy benefits of the sliced OPT-based method in various numerical experiments. In particular, we show an application of our proposed Sliced-OPT in noisy point cloud registration.

♻ ☆ Fast expansion into harmonics on the disk: a steerable basis with fast radial convolutions

Nicholas F. Marshall, Oscar Mickelin, Amit Singer

We present a fast and numerically accurate method for expanding digitized $L \times L$ images representing functions on $[-1,1]^2$ supported on the disk $\{x \in \mathbb{R}^2 : |x|<1\}$ in the harmonics (Dirichlet Laplacian eigenfunctions) on the disk. Our method, which we refer to as the Fast Disk Harmonics Transform (FDHT), runs in $O(L^2 \log L)$ operations. This basis is also known as the Fourier-Bessel basis, and it has several computational advantages: it is orthogonal, ordered by frequency, and steerable in the sense that images expanded in the basis can be rotated by applying a diagonal transform to the coefficients. Moreover, we show that convolution with radial functions can also be efficiently computed by applying a diagonal transform to the coefficients.

comment: 26 pages, 5 figures, 1 table

♻ ☆ Convergence of Invariant Graph Networks

Chen Cai, Yusu Wang

Although theoretical properties such as expressive power and over-smoothing of graph neural networks (GNN) have been extensively studied recently, its convergence property is a relatively new direction. In this paper, we investigate the convergence of one powerful GNN, Invariant Graph Network (IGN) over graphs sampled from graphons. We first prove the stability of linear layers for general $k$-IGN (of order $k$) based on a novel interpretation of linear equivariant layers. Building upon this result, we prove the convergence of $k$-IGN under the model of \citet{ruiz2020graphon}, where we access the edge weight but the convergence error is measured for graphon inputs. Under the more natural (and more challenging) setting of \citet{keriven2020convergence} where one can only access 0-1 adjacency matrix sampled according to edge probability, we first show a negative result that the convergence of any IGN is not possible. We then obtain the convergence of a subset of IGNs, denoted as IGN-small, after the edge probability estimation. We show that IGN-small still contains function class rich enough that can approximate spectral GNNs arbitrarily well. Lastly, we perform experiments on various graphon models to verify our statements.

comment: 29 pages, 11 figures

♻ ☆ Multiple Imputation with Neural Network Gaussian Process for High-dimensional Incomplete Data

Zongyu Dai, Zhiqi Bu, Qi Long

Missing data are ubiquitous in real world applications and, if not adequately handled, may lead to the loss of information and biased findings in downstream analysis. Particularly, high-dimensional incomplete data with a moderate sample size, such as analysis of multi-omics data, present daunting challenges. Imputation is arguably the most popular method for handling missing data, though existing imputation methods have a number of limitations. Single imputation methods such as matrix completion methods do not adequately account for imputation uncertainty and hence would yield improper statistical inference. In contrast, multiple imputation (MI) methods allow for proper inference but existing methods do not perform well in high-dimensional settings. Our work aims to address these significant methodological gaps, leveraging recent advances in neural network Gaussian process (NNGP) from a Bayesian viewpoint. We propose two NNGP-based MI methods, namely MI-NNGP, that can apply multiple imputations for missing values from a joint (posterior predictive) distribution. The MI-NNGP methods are shown to significantly outperform existing state-of-the-art methods on synthetic and real datasets, in terms of imputation error, statistical inference, robustness to missing rates, and computation costs, under three missing data mechanisms, MCAR, MAR, and MNAR.

♻ ☆ Low Variance Off-policy Evaluation with State-based Importance Sampling

David M. Bossens, Philip Thomas

In off-policy reinforcement learning, a behaviour policy performs exploratory interactions with the environment to obtain state-action-reward samples which are then used to learn a target policy that optimises the expected return. This leads to a problem of off-policy evaluation, where one needs to evaluate the target policy from samples collected by the often unrelated behaviour policy. Importance sampling is a traditional statistical technique that is often applied to off-policy evaluation. While importance sampling estimators are unbiased, their variance increases exponentially with the horizon of the decision process due to computing the importance weight as a product of action probability ratios, yielding estimates with low accuracy for domains involving long-term planning. This paper proposes state-based importance sampling (SIS), which drops the action probability ratios of sub-trajectories with "neglible states" -- roughly speaking, those for which the chosen actions have no impact on the return estimate -- from the computation of the importance weight. Theoretical results show that this results in a reduction of the exponent in the variance upper bound as well as improving the mean squared error. An automated search algorithm based on covariance testing is proposed to identify a negligible state set which has minimal MSE when performing state-based importance sampling. Experiments are conducted on a lift domain, which include "lift states" where the action has no impact on the following state and reward. The results demonstrate that using the search algorithm, SIS yields reduced variance and improved accuracy compared to traditional importance sampling, per-decision importance sampling, and incremental importance sampling.

♻ ☆ End-to-End Learned Early Classification of Time Series for In-Season Crop Type Mapping SP

Marc Rußwurm, Nicolas Courty, Rémi Emonet, Sébastien Lefèvre, Devis Tuia, Romain Tavenard

Remote sensing satellites capture the cyclic dynamics of our Planet in regular time intervals recorded in satellite time series data. End-to-end trained deep learning models use this time series data to make predictions at a large scale, for instance, to produce up-to-date crop cover maps. Most time series classification approaches focus on the accuracy of predictions. However, the earliness of the prediction is also of great importance since coming to an early decision can make a crucial difference in time-sensitive applications. In this work, we present an End-to-End Learned Early Classification of Time Series (ELECTS) model that estimates a classification score and a probability of whether sufficient data has been observed to come to an early and still accurate decision. ELECTS is modular: any deep time series classification model can adopt the ELECTS conceptual idea by adding a second prediction head that outputs a probability of stopping the classification. The ELECTS loss function then optimizes the overall model on a balanced objective of earliness and accuracy. Our experiments on four crop classification datasets from Europe and Africa show that ELECTS allows reaching state-of-the-art accuracy while reducing the quantity of data massively to be downloaded, stored, and processed. The source code is available at https://github.com/marccoru/elects.

comment: accepted for publication in ISPRS Journal of Photogrammetry and Remote Sensing

♻ ☆ A Simple Way to Learn Metrics Between Attributed Graphs

Yacouba Kaloga, Pierre Borgnat, Amaury Habrard

The choice of good distances and similarity measures between objects is important for many machine learning methods. Therefore, many metric learning algorithms have been developed in recent years, mainly for Euclidean data in order to improve performance of classification or clustering methods. However, due to difficulties in establishing computable, efficient and differentiable distances between attributed graphs, few metric learning algorithms adapted to graphs have been developed despite the strong interest of the community. In this paper, we address this issue by proposing a new Simple Graph Metric Learning - SGML - model with few trainable parameters based on Simple Graph Convolutional Neural Networks - SGCN - and elements of Optimal Transport theory. This model allows us to build an appropriate distance from a database of labeled (attributed) graphs to improve the performance of simple classification algorithms such as $k$-NN. This distance can be quickly trained while maintaining good performances as illustrated by the experimental study presented in this paper.

♻ ☆ Black-box tests for algorithmic stability

Byol Kim, Rina Foygel Barber

Algorithmic stability is a concept from learning theory that expresses the degree to which changes to the input data (e.g., removal of a single data point) may affect the outputs of a regression algorithm. Knowing an algorithm's stability properties is often useful for many downstream applications -- for example, stability is known to lead to desirable generalization properties and predictive inference guarantees. However, many modern algorithms currently used in practice are too complex for a theoretical analysis of their stability properties, and thus we can only attempt to establish these properties through an empirical exploration of the algorithm's behavior on various data sets. In this work, we lay out a formal statistical framework for this kind of "black-box testing" without any assumptions on the algorithm or the data distribution and establish fundamental bounds on the ability of any black-box test to identify algorithmic stability.

comment: 37 pages. Minor edits to match the journal-submitted version

♻ ☆ An Algorithm for Routing Vectors in Sequences

Franz A. Heinsen

We propose a routing algorithm that takes a sequence of vectors and computes a new sequence with specified length and vector size. Each output vector maximizes "bang per bit," the difference between a net benefit to use and net cost to ignore data, by better predicting the input vectors. We describe output vectors as geometric objects, as latent variables that assign credit, as query states in a model of associative memory, and as agents in a model of a Society of Mind. We implement the algorithm with optimizations that reduce parameter count, computation, and memory use by orders of magnitude, enabling us to route sequences of greater length than previously possible. We evaluate our implementation on natural language and visual classification tasks, obtaining competitive or state-of-the-art accuracy and end-to-end credit assignments that are interpretable.

comment: Source code and instructions for replicating our results are online at https://github.com/glassroom/heinsen_routing

♻ ☆ CSI: Contrastive Data Stratification for Interaction Prediction and its Application to Compound-Protein Interaction Prediction

Apurva Kalia, Dilip Krishnan, Soha Hassoun

Accurately predicting the likelihood of interaction between two objects (compound-protein sequence, user-item, author-paper, etc.) is a fundamental problem in Computer Science. Current deep-learning models rely on learning accurate representations of the interacting objects. Importantly, relationships between the interacting objects, or features of the interaction, offer an opportunity to partition the data to create multi-views of the interacting objects. The resulting congruent and non-congruent views can then be exploited via contrastive learning techniques to learn enhanced representations of the objects.

comment: 11 pages, submitted to BioInformatics

☆ Attention-Aware Anime Line Drawing Colorization

Yu Cao, Hao Tian, P. Y. Mok

Automatic colorization of anime line drawing has attracted much attention in recent years since it can substantially benefit the animation industry. User-hint based methods are the mainstream approach for line drawing colorization, while reference-based methods offer a more intuitive approach. Nevertheless, although reference-based methods can improve feature aggregation of the reference image and the line drawing, the colorization results are not compelling in terms of color consistency or semantic correspondence. In this paper, we introduce an attention-based model for anime line drawing colorization, in which a channel-wise and spatial-wise Convolutional Attention module is used to improve the ability of the encoder for feature extraction and key area perception, and a Stop-Gradient Attention module with cross-attention and self-attention is used to tackle the cross-domain long-range dependency problem. Extensive experiments show that our method outperforms other SOTA methods, with more accurate line structure and semantic color information.

☆ RECAP: Retrieval Augmented Music Captioner

Zihao He, Weituo Hao, Xuchen Song

With the prevalence of stream media platforms serving music search and recommendation, interpreting music by understanding audio and lyrics interactively has become an important and challenging task. However, many previous works focus on refining individual components of encoder-decoder architecture mapping music to caption tokens, ignoring the potential usage of audio and lyrics correspondence. In this paper, we propose to explicitly learn the multi-modal alignment with retrieval augmentation by contrastive learning. By learning audio-lyrics correspondence, the model is guided to learn better cross-modal attention weights, thus generating high-quality caption words. We provide both theoretical and empirical results that demonstrate the advantage of the proposed method.

☆ From Images to Textual Prompts: Zero-shot VQA with Frozen Large Language Models

Jiaxian Guo, Junnan Li, Dongxu Li, Anthony Meng Huat Tiong, Boyang Li, Dacheng Tao, Steven C. H. Hoi

Large language models (LLMs) have demonstrated excellent zero-shot generalization to new language tasks. However, effective utilization of LLMs for zero-shot visual question-answering (VQA) remains challenging, primarily due to the modality disconnection and task disconnection between LLM and VQA task. End-to-end training on vision and language data may bridge the disconnections, but is inflexible and computationally expensive. To address this issue, we propose \emph{Img2Prompt}, a plug-and-play module that provides the prompts that can bridge the aforementioned modality and task disconnections, so that LLMs can perform zero-shot VQA tasks without end-to-end training. In order to provide such prompts, we further employ LLM-agnostic models to provide prompts that can describe image content and self-constructed question-answer pairs, which can effectively guide LLM to perform zero-shot VQA tasks. Img2Prompt offers the following benefits: 1) It can flexibly work with various LLMs to perform VQA. 2)~Without the needing of end-to-end training, it significantly reduces the cost of deploying LLM for zero-shot VQA tasks. 3) It achieves comparable or better performance than methods relying on end-to-end training. For example, we outperform Flamingo~\cite{Deepmind:Flamingo2022} by 5.6\% on VQAv2. On the challenging A-OKVQA dataset, our method even outperforms few-shot methods by as much as 20\%.

☆ Cattle Detection Occlusion Problem

Aparna Mendu, Bhavya Sehgal, Vaishnavi Mendu

The management of cattle over a huge area is still a challenging problem in the farming sector. With evolution in technology, Unmanned aerial vehicles (UAVs) with consumer level digital cameras are becoming a popular alternative to manual animal censuses for livestock estimation since they are less risky and expensive.This paper evaluated and compared the cutting-edge object detection algorithms, YOLOv7,RetinaNet with ResNet50 backbone, RetinaNet with EfficientNet and mask RCNN. It aims to improve the occlusion problem that is to detect hidden cattle from a huge dataset captured by drones using deep learning algorithms for accurate cattle detection. Experimental results showed YOLOv7 was superior with precision of 0.612 when compared to the other two algorithms. The proposed method proved superior to the usual competing algorithms for cow face detection, especially in very difficult cases.

comment: arXiv admin note: text overlap with arXiv:1701.01611 by other authors

☆ Advanced Baseline for 3D Human Pose Estimation: A Two-Stage Approach

Zichen Gui, Jungang Luo

Human pose estimation has been widely applied in various industries. While recent decades have witnessed the introduction of many advanced two-dimensional (2D) human pose estimation solutions, three-dimensional (3D) human pose estimation is still an active research field in computer vision. Generally speaking, 3D human pose estimation methods can be divided into two categories: single-stage and two-stage. In this paper, we focused on the 2D-to-3D lifting process in the two-stage methods and proposed a more advanced baseline model for 3D human pose estimation, based on the existing solutions. Our improvements include optimization of machine learning models and multiple parameters, as well as introduction of a weighted loss to the training model. Finally, we used the Human3.6M benchmark to test the final performance and it did produce satisfactory results.

♻ ☆ MMDialog: A Large-scale Multi-turn Dialogue Dataset Towards Multi-modal Open-domain Conversation

Jiazhan Feng, Qingfeng Sun, Can Xu, Pu Zhao, Yaming Yang, Chongyang Tao, Dongyan Zhao, Qingwei Lin

Responding with multi-modal content has been recognized as an essential capability for an intelligent conversational agent. In this paper, we introduce the MMDialog dataset to better facilitate multi-modal conversation. MMDialog is composed of a curated set of 1.08 million real-world dialogues with 1.53 million unique images across 4,184 topics. MMDialog has two main and unique advantages. First, it is the largest multi-modal conversation dataset by the number of dialogues by 88x. Second, it contains massive topics to generalize the open-domain. To build engaging dialogue system with this dataset, we propose and normalize two response producing tasks based on retrieval and generative scenarios. In addition, we build two baselines for above tasks with state-of-the-art techniques and report their experimental performance. We also propose a novel evaluation metric MM-Relevance to measure the multi-modal responses. Our dataset and scripts are available in https://github.com/victorsungo/MMDialog.

☆ Does unsupervised grammar induction need pixels?

Boyi Li, Rodolfo Corona, Karttikeya Mangalam, Catherine Chen, Daniel Flaherty, Serge Belongie, Kilian Q. Weinberger, Jitendra Malik, Trevor Darrell, Dan Klein

Are extralinguistic signals such as image pixels crucial for inducing constituency grammars? While past work has shown substantial gains from multimodal cues, we investigate whether such gains persist in the presence of rich information from large language models (LLMs). We find that our approach, LLM-based C-PCFG (LC-PCFG), outperforms previous multi-modal methods on the task of unsupervised constituency parsing, achieving state-of-the-art performance on a variety of datasets. Moreover, LC-PCFG results in an over 50% reduction in parameter count, and speedups in training time of 1.7x for image-aided models and more than 5x for video-aided models, respectively. These results challenge the notion that extralinguistic signals such as image pixels are needed for unsupervised grammar induction, and point to the need for better text-only baselines in evaluating the need of multi-modality for the task.

☆ Debiasing NLP Models Without Demographic Information

Hadas Orgad, Yonatan Belinkov

Models trained from real-world data tend to imitate and amplify social biases. Although there are many methods suggested to mitigate biases, they require a preliminary information on the types of biases that should be mitigated (e.g., gender or racial bias) and the social groups associated with each data sample. In this work, we propose a debiasing method that operates without any prior knowledge of the demographics in the dataset, detecting biased examples based on an auxiliary model that predicts the main model's success and down-weights them during the training process. Results on racial and gender bias demonstrate that it is possible to mitigate social biases without having to use a costly demographic annotation process.

☆ Parsel: A Unified Natural Language Framework for Algorithmic Reasoning

Eric Zelikman, Qian Huang, Gabriel Poesia, Noah D. Goodman, Nick Haber

Despite recent success in large language model (LLM) reasoning, LLMs still struggle with hierarchical multi-step reasoning like generating complex programs. In these cases, humans often start with a high-level algorithmic design and implement each part gradually. We introduce Parsel, a framework enabling automatic implementation and validation of complex algorithms with code LLMs, based on hierarchical function descriptions in natural language. Parsel can be used across domains requiring hierarchical reasoning, e.g. code synthesis, theorem proving, and robotic planning. We demonstrate Parsel's capabilities by using it to generate complex programs that cannot currently be automatically implemented from one description and backtranslating Python programs in the APPS dataset. Beyond modeling capabilities, Parsel allows problem-solving with high-level algorithmic designs, benefiting both students and professional programmers.

☆ Character-Aware Models Improve Visual Text Rendering

Rosanne Liu, Dan Garrette, Chitwan Saharia, William Chan, Adam Roberts, Sharan Narang, Irina Blok, RJ Mical, Mohammad Norouzi, Noah Constant

Current image generation models struggle to reliably produce well-formed visual text. In this paper, we investigate a key contributing factor: popular text-to-image models lack character-level input features, making it much harder to predict a word's visual makeup as a series of glyphs. To quantify the extent of this effect, we conduct a series of controlled experiments comparing character-aware vs. character-blind text encoders. In the text-only domain, we find that character-aware models provide large gains on a novel spelling task (WikiSpell). Transferring these learnings onto the visual domain, we train a suite of image generation models, and show that character-aware variants outperform their character-blind counterparts across a range of novel text rendering tasks (our DrawText benchmark). Our models set a much higher state-of-the-art on visual spelling, with 30+ point accuracy gains over competitors on rare words, despite training on far fewer examples.

☆ Self-Instruct: Aligning Language Model with Self Generated Instructions

Yizhong Wang, Yeganeh Kordi, Swaroop Mishra, Alisa Liu, Noah A. Smith, Daniel Khashabi, Hannaneh Hajishirzi

Large "instruction-tuned" language models (finetuned to respond to instructions) have demonstrated a remarkable ability to generalize zero-shot to new tasks. Nevertheless, they depend heavily on human-written instruction data that is limited in quantity, diversity, and creativity, therefore hindering the generality of the tuned model. We introduce Self-Instruct, a framework for improving the instruction-following capabilities of pretrained language models by bootstrapping off its own generations. Our pipeline generates instruction, input, and output samples from a language model, then prunes them before using them to finetune the original model. Applying our method to vanilla GPT3, we demonstrate a 33% absolute improvement over the original model on Super-NaturalInstructions, on par with the performance of InstructGPT_001, which is trained with private user data and human annotations. For further evaluation, we curate a set of expert-written instructions for novel tasks, and show through human evaluation that tuning GPT3 with Self-Instruct outperforms using existing public instruction datasets by a large margin, leaving only a 5% absolute gap behind InstructGPT_001. Self-Instruct provides an almost annotation-free method for aligning pre-trained language models with instructions, and we release our large synthetic dataset to facilitate future studies on instruction tuning.

comment: work in progress

☆ Why Can GPT Learn In-Context? Language Models Secretly Perform Gradient Descent as Meta Optimizers

Damai Dai, Yutao Sun, Li Dong, Yaru Hao, Zhifang Sui, Furu Wei

Large pretrained language models have shown surprising In-Context Learning (ICL) ability. With a few demonstration input-label pairs, they can predict the label for an unseen input without additional parameter updates. Despite the great success in performance, the working mechanism of ICL still remains an open problem. In order to better understand how ICL works, this paper explains language models as meta optimizers and understands ICL as a kind of implicit finetuning. Theoretically, we figure out that the Transformer attention has a dual form of gradient descent based optimization. On top of it, we understand ICL as follows: GPT first produces meta gradients according to the demonstration examples, and then these meta gradients are applied to the original GPT to build an ICL model. Experimentally, we comprehensively compare the behavior of ICL and explicit finetuning based on real tasks to provide empirical evidence that supports our understanding. The results prove that ICL behaves similarly to explicit finetuning at the prediction level, the representation level, and the attention behavior level. Further, inspired by our understanding of meta optimization, we design a momentum-based attention by analogy with the momentum-based gradient descent algorithm. Its consistently better performance over vanilla attention supports our understanding again from another aspect, and more importantly, it shows the potential to utilize our understanding for future model designing.

☆ On-the-fly Denoising for Data Augmentation in Natural Language Understanding

Tianqing Fang, Wenxuan Zhou, Fangyu Liu, Hongming Zhang, Yangqiu Song, Muhao Chen

Data Augmentation (DA) is frequently used to automatically provide additional training data without extra human annotation. However, data augmentation may introduce noisy data that impairs training. To guarantee the quality of augmented data, existing methods either assume no noise exists in the augmented data and adopt consistency training or use simple heuristics such as training loss and diversity constraints to filter out ``noisy'' data. However, those filtered examples may still contain useful information, and dropping them completely causes loss of supervision signals. In this paper, based on the assumption that the original dataset is cleaner than the augmented data, we propose an on-the-fly denoising technique for data augmentation that learns from soft augmented labels provided by an organic teacher model trained on the cleaner original data. A simple self-regularization module is applied to force the model prediction to be consistent across two distinct dropouts to further prevent overfitting on noisy labels. Our method can be applied to augmentation techniques in general and can consistently improve the performance on both text classification and question-answering tasks.

comment: 14 pages

☆ DialGuide: Aligning Dialogue Model Behavior with Developer Guidelines

Prakhar Gupta, Yang Liu, Di Jin, Behnam Hedayatnia, Spandana Gella, Sijia Liu, Patrick Lange, Julia Hirschberg, Dilek Hakkani-Tur

Dialogue models are able to generate coherent and fluent responses, but they can still be challenging to control and may produce non-engaging, unsafe results. This unpredictability diminishes user trust and can hinder the use of the models in the real world. To address this, we introduce DialGuide, a novel framework for controlling dialogue model behavior using natural language rules, or guidelines. These guidelines provide information about the context they are applicable to and what should be included in the response, allowing the models to generate responses that are more closely aligned with the developer's expectations and intent. We evaluate DialGuide on three tasks in open-domain dialogue response generation: guideline selection, response generation, and response entailment verification. Our dataset contains 10,737 positive and 15,467 negative dialogue context-response-guideline triplets across two domains - chit-chat and safety. We provide baseline models for the tasks and benchmark their performance. We also demonstrate that DialGuide is effective in the dialogue safety domain, producing safe and engaging responses that follow developer guidelines.

☆ PairReranker: Pairwise Reranking for Natural Language Generation

Dongfu Jiang, Bill Yuchen Lin, Xiang Ren

Pre-trained language models have been successful in natural language generation (NLG) tasks. While various decoding methods have been employed, they often produce suboptimal results. We first present an empirical analysis of three NLG tasks: summarization, machine translation, and constrained text generation. We found that selecting the best output from the results of multiple decoding methods can significantly improve performance. To further improve reranking for NLG tasks, we proposed a novel method, \textsc{PairReranker}, which uses a single encoder and a pairwise loss function to jointly encode a source input and a pair of candidates and compare them. Experiments on three NLG tasks demonstrated the effectiveness and flexibility of \textsc{PairReranker}, showing strong results, compared with previous baselines. In addition, our \textsc{PairReranker} can generalize to significantly improve GPT-3 (text-davinci-003) results (e.g., 24.55\% on CommonGen and 11.35\% on WMT18 zh-en), even though our rerankers are not trained with any GPT-3 candidates.

comment: We will release our code and data at https://inklab.usc.edu/PairReranker

☆ A Length-Extrapolatable Transformer

Yutao Sun, Li Dong, Barun Patra, Shuming Ma, Shaohan Huang, Alon Benhaim, Vishrav Chaudhary, Xia Song, Furu Wei

Position modeling plays a critical role in Transformers. In this paper, we focus on length extrapolation, i.e., training on short texts while evaluating longer sequences. We define attention resolution as an indicator of extrapolation. Then we propose two designs to improve the above metric of Transformers. Specifically, we introduce a relative position embedding to explicitly maximize attention resolution. Moreover, we use blockwise causal attention during inference for better resolution. We evaluate different Transformer variants with language modeling. Experimental results show that our model achieves strong performance in both interpolation and extrapolation settings. The code will be available at https://aka.ms/LeX-Transformer.

comment: 9 pages

☆ Lego-MT: Towards Detachable Models in Massively Multilingual Machine Translation

Fei Yuan, Yinquan Lu, WenHao Zhu, Lingpeng Kong, Lei Li, Jingjing Xu

Traditional multilingual neural machine translation (MNMT) uses a single model to translate all directions. However, with the increasing scale of language pairs, simply using a single model for massive MNMT brings new challenges: parameter tension and large computations. In this paper, we revisit multi-way structures by assigning an individual branch for each language (group). Despite being a simple architecture, it is challenging to train de-centralized models due to the lack of constraints to align representations from all languages. We propose a localized training recipe to map different branches into a unified space, resulting in an efficient detachable model, Lego-MT. For a fair comparison, we collect data from OPUS and build the first large-scale open-source translation benchmark covering 7 language-centric data, each containing 445 language pairs. Experiments show that Lego-MT (1.2B) brings gains of more than 4 BLEU while outperforming M2M-100 (12B) (We will public all training data, models, and checkpoints)

☆ Cross-modal Attention Congruence Regularization for Vision-Language Relation Alignment

Rohan Pandey, Rulin Shao, Paul Pu Liang, Ruslan Salakhutdinov, Louis-Philippe Morency

Despite recent progress towards scaling up multimodal vision-language models, these models are still known to struggle on compositional generalization benchmarks such as Winoground. We find that a critical component lacking from current vision-language models is relation-level alignment: the ability to match directional semantic relations in text (e.g., "mug in grass") with spatial relationships in the image (e.g., the position of the mug relative to the grass). To tackle this problem, we show that relation alignment can be enforced by encouraging the directed language attention from 'mug' to 'grass' (capturing the semantic relation 'in') to match the directed visual attention from the mug to the grass. Tokens and their corresponding objects are softly identified using the cross-modal attention. We prove that this notion of soft relation alignment is equivalent to enforcing congruence between vision and language attention matrices under a 'change of basis' provided by the cross-modal attention matrix. Intuitively, our approach projects visual attention into the language attention space to calculate its divergence from the actual language attention, and vice versa. We apply our Cross-modal Attention Congruence Regularization (CACR) loss to UNITER and improve on the state-of-the-art approach to Winoground.

☆ T-Projection: High Quality Annotation Projection for Sequence Labeling Tasks

Iker García-Ferrero, Rodrigo Agerri, German Rigau

In the absence of readily available labeled data for a given task and language, annotation projection has been proposed as one of the possible strategies to automatically generate annotated data which may then be used to train supervised systems. Annotation projection has often been formulated as the task of projecting, on parallel corpora, some labels from a source into a target language. In this paper we present T-Projection, a new approach for annotation projection that leverages large pretrained text2text language models and state-of-the-art machine translation technology. T-Projection decomposes the label projection task into two subtasks: (i) The candidate generation step, in which a set of projection candidates using a multilingual T5 model is generated and, (ii) the candidate selection step, in which the candidates are ranked based on translation probabilities. We evaluate our method in three downstream tasks and five different languages. Our results show that T-projection improves the average F1 score of previous methods by more than 8 points.

☆ Semantically-informed Hierarchical Event Modeling

Shubhashis Roy Dipta, Mehdi Rezaee, Francis Feraro

Prior work has shown that coupling sequential latent variable models with semantic ontological knowledge can improve the representational capabilities of event modeling approaches. In this work, we present a novel, doubly hierarchical, semi-supervised event modeling framework that provides structural hierarchy while also accounting for ontological hierarchy. Our approach consists of multiple layers of structured latent variables, where each successive layer compresses and abstracts the previous layers. We guide this compression through the injection of structured ontological knowledge that is defined at the type level of events: importantly, our model allows for partial injection of semantic knowledge and it does not depend on observing instances at any particular level of the semantic ontology. Across two different datasets and four different evaluation metrics, we demonstrate that our approach is able to out-perform the previous state-of-the-art approaches, demonstrating the benefits of structured and semantic hierarchical knowledge for event modeling.

☆ DimonGen: Diversified Generative Commonsense Reasoning for Explaining Concept Relationships

Chenzhengyi Liu, Jie Huang, Kerui Zhu, Kevin Chen-Chuan Chang

In this paper, we propose DimonGen, which aims to generate diverse sentences describing concept relationships in various everyday scenarios. To support this, we create a benchmark dataset for this task by adapting the existing CommonGen dataset and propose a two-stage model called MoREE (Mixture of Retrieval-Enhanced Experts) to generate the target sentences. MoREE consists of a mixture of retriever models that retrieve diverse context sentences related to the given concepts, and a mixture of generator models that generate diverse sentences based on the retrieved contexts. We conduct experiments on the DimonGen task and show that MoREE outperforms strong baselines in terms of both the quality and diversity of the generated sentences. Our results demonstrate that MoREE is able to generate diverse sentences that reflect different relationships between concepts, leading to a comprehensive understanding of concept relationships.

☆ Pretraining Without Attention

Junxiong Wang, Jing Nathan Yan, Albert Gu, Alexander M. Rush

Transformers have been essential to pretraining success in NLP. Other architectures have been used, but require attention layers to match benchmark accuracy. This work explores pretraining without attention. We test recently developed routing layers based on state-space models (SSM) and model architectures based on multiplicative gating. Used together these modeling choices have a large impact on pretraining accuracy. Empirically the proposed Bidirectional Gated SSM (BiGS) replicates BERT pretraining results without attention and can be extended to long-form pretraining of 4096 tokens without approximation.

☆ Detoxifying Text with MaRCo: Controllable Revision with Experts and Anti-Experts

Skyler Hallinan, Alisa Liu, Yejin Choi, Maarten Sap

Text detoxification has the potential to mitigate the harms of toxicity by rephrasing text to remove offensive meaning, but subtle toxicity remains challenging to tackle. We introduce MaRCo, a detoxification algorithm that combines controllable generation and text rewriting methods using a Product of Experts with autoencoder language models (LMs). MaRCo uses likelihoods under a non-toxic LM (expert) and a toxic LM (anti-expert) to find candidate words to mask and potentially replace. We evaluate our method on several subtle toxicity and microaggressions datasets, and show that it not only outperforms baselines on automatic metrics, but MaRCo's rewrites are preferred 2.1 $\times$ more in human evaluation. Its applicability to instances of subtle toxicity is especially promising, demonstrating a path forward for addressing increasingly elusive online hate.

☆ Toward Human Readable Prompt Tuning: Kubrick's The Shining is a good movie, and a good prompt too?

Weijia Shi, Xiaochuang Han, Hila Gonen, Ari Holtzman, Yulia Tsvetkov, Luke Zettlemoyer

Large language models can perform new tasks in a zero-shot fashion, given natural language prompts that specify the desired behavior. Such prompts are typically hand engineered, but can also be learned with gradient-based methods from labeled data. However, it is underexplored what factors make the prompts effective, especially when the prompts are natural language. In this paper, we investigate common attributes shared by effective prompts. We first propose a human readable prompt tuning method (F LUENT P ROMPT) based on Langevin dynamics that incorporates a fluency constraint to find a diverse distribution of effective and fluent prompts. Our analysis reveals that effective prompts are topically related to the task domain and calibrate the prior probability of label words. Based on these findings, we also propose a method for generating prompts using only unlabeled data, outperforming strong baselines by an average of 7.0% accuracy across three tasks.

☆ Does CLIP Bind Concepts? Probing Compositionality in Large Image Models

Martha Lewis, Qinan Yu, Jack Merullo, Ellie Pavlick

Large-scale models combining text and images have made incredible progress in recent years. However, they can still fail at tasks requiring compositional knowledge, such as correctly picking out a red cube from a picture of multiple shapes. We examine the ability of CLIP (Radford et al., 2021), to caption images requiring compositional knowledge. We implement five compositional language models to probe the kinds of structure that CLIP may be using, and develop a novel training algorithm, Compositional Skipgram for Images (CoSI), to train these models. We look at performance in attribute-based tasks, requiring the identification of a particular combination of attribute and object (such as "red cube"), and in relational settings, where the spatial relation between two shapes (such as "cube behind sphere") must be identified. We find that in some conditions, CLIP is able to learn attribute-object labellings, and to generalize to unseen attribute-object combinations. However, we also see evidence that CLIP is not able to bind features together reliably. Moreover, CLIP is not able to reliably learn relations between objects, whereas some compositional models are able to learn these perfectly. Of the five models we developed, none were able to generalize to unseen relations.

☆ Measure More, Question More: Experimental Studies on Transformer-based Language Models and Complement Coercion

Yuling Gu

Transformer-based language models have shown strong performance on an array of natural language understanding tasks. However, the question of how these models react to implicit meaning has been largely unexplored. We investigate this using the complement coercion phenomenon, which involves sentences like "The student finished the book about sailing" where the action "reading" is implicit. We compare LMs' surprisal estimates at various critical sentence regions in sentences with and without implicit meaning. Effects associated with recovering implicit meaning were found at a critical region other than where sentences minimally differ. We then use follow-up experiments to factor out potential confounds, revealing different perspectives that offer a richer and more accurate picture.

☆ A Survey of Deep Learning for Mathematical Reasoning

Pan Lu, Liang Qiu, Wenhao Yu, Sean Welleck, Kai-Wei Chang

Mathematical reasoning is a fundamental aspect of human intelligence and is applicable in various fields, including science, engineering, finance, and everyday life. The development of artificial intelligence (AI) systems capable of solving math problems and proving theorems has garnered significant interest in the fields of machine learning and natural language processing. For example, mathematics serves as a testbed for aspects of reasoning that are challenging for powerful deep learning models, driving new algorithmic and modeling advances. On the other hand, recent advances in large-scale neural language models have opened up new benchmarks and opportunities to use deep learning for mathematical reasoning. In this survey paper, we review the key tasks, datasets, and methods at the intersection of mathematical reasoning and deep learning over the past decade. We also evaluate existing benchmarks and methods, and discuss future research directions in this domain.

comment: 24 pages, 2 figures, 8 tables. The repository is available at https://github.com/lupantech/dl4math

☆ DISCO: Distilling Phrasal Counterfactuals with Large Language Models

Zeming Chen, Qiyue Gao, Kyle Richardson, Antoine Bosselut, Ashish Sabharwal

Recent methods demonstrate that data augmentation using counterfactual knowledge can teach models the causal structure of a task, leading to robust and generalizable models. However, such counterfactual data often has a limited scale and diversity if crowdsourced and is computationally expensive to extend to new perturbation types if generated using supervised methods. To address this, we introduce a new framework called DISCO for automatically generating high-quality counterfactual data at scale. DISCO engineers prompts to generate phrasal perturbations with a large general language model. Then, a task-specific teacher model filters the generation to distill high-quality counterfactual data. We show that learning with this counterfactual data yields a comparatively small student model that is 6% (absolute) more robust and generalizes 5% better across distributions than baselines on various challenging evaluations. This model is also 15% more sensitive in differentiating original and counterfactual examples, on three evaluation sets written by human workers and via human-AI collaboration.

comment: work in progress

☆ Is GPT-3 a Psychopath? Evaluating Large Language Models from a Psychological Perspective

Xingxuan Li, Yutong Li, Linlin Liu, Lidong Bing, Shafiq Joty

Are large language models (LLMs) like GPT-3 psychologically safe? In this work, we design unbiased prompts to evaluate LLMs systematically from a psychological perspective. Firstly, we test the personality traits of three different LLMs with Short Dark Triad (SD-3) and Big Five Inventory (BFI). We find all of them show higher scores on SD-3 than the human average, indicating a relatively darker personality. Furthermore, LLMs like InstructGPT and FLAN-T5, which are fine-tuned with safety metrics, do not necessarily have more positive personalities. They score higher on Machiavellianism and Narcissism than GPT-3. Secondly, we test the LLMs in GPT-3 series on well-being tests to study the impact of fine-tuning with more training data. Interestingly, we observe a continuous increase in well-being scores from GPT-3 to InstructGPT. Following the observations, we show that instruction-finetune FLAN-T5 with positive answers in BFI can effectively improve the model from a psychological perspective. Finally, we call on the community to evaluate and improve LLMs' safety systematically instead of at the sentence level only.

☆ HYRR: Hybrid Infused Reranking for Passage Retrieval

Jing Lu, Keith Hall, Ji Ma, Jianmo Ni

We present Hybrid Infused Reranking for Passages Retrieval (HYRR), a framework for training rerankers based on a hybrid of BM25 and neural retrieval models. Retrievers based on hybrid models have been shown to outperform both BM25 and neural models alone. Our approach exploits this improved performance when training a reranker, leading to a robust reranking model. The reranker, a cross-attention neural model, is shown to be robust to different first-stage retrieval systems, achieving better performance than rerankers simply trained upon the first-stage retrievers in the multi-stage systems. We present evaluations on a supervised passage retrieval task using MS MARCO and zero-shot retrieval tasks using BEIR. The empirical results show strong performance on both evaluations.

☆ Exploring the Challenges of Open Domain Multi-Document Summarization

John Giorgi, Luca Soldaini, Bo Wang, Gary Bader, Kyle Lo, Lucy Lu Wang, Arman Cohan

Multi-document summarization (MDS) has traditionally been studied assuming a set of ground-truth topic-related input documents is provided. In practice, the input document set is unlikely to be available a priori and would need to be retrieved based on an information need, a setting we call open-domain MDS. We experiment with current state-of-the-art retrieval and summarization models on several popular MDS datasets extended to the open-domain setting. We find that existing summarizers suffer large reductions in performance when applied as-is to this more realistic task, though training summarizers with retrieved inputs can reduce their sensitivity retrieval errors. To further probe these findings, we conduct perturbation experiments on summarizer inputs to study the impact of different types of document retrieval errors. Based on our results, we provide practical guidelines to help facilitate a shift to open-domain MDS. We release our code and experimental results alongside all data or model artifacts created during our investigation.

comment: Work in progress

☆ SLUE Phase-2: A Benchmark Suite of Diverse Spoken Language Understanding Tasks

Suwon Shon, Siddhant Arora, Chyi-Jiunn Lin, Ankita Pasad, Felix Wu, Roshan Sharma, Wei-Lun Wu, Hung-Yi Lee, Karen Livescu, Shinji Watanabe

Spoken language understanding (SLU) tasks have been studied for many decades in the speech research community, but have not received as much attention as lower-level tasks like speech and speaker recognition. In particular, there are not nearly as many SLU task benchmarks, and many of the existing ones use data that is not freely available to all researchers. Recent work has begun to introduce such benchmark datasets for several tasks. In this work, we introduce several new annotated SLU benchmark tasks based on freely available speech data, which complement existing benchmarks and address gaps in the SLU evaluation landscape. We contribute four tasks: question answering and summarization involve inference over longer speech sequences; named entity localization addresses the speech-specific task of locating the targeted content in the signal; dialog act classification identifies the function of a given speech utterance. We follow the blueprint of the Spoken Language Understanding Evaluation (SLUE) benchmark suite. In order to facilitate the development of SLU models that leverage the success of pre-trained speech representations, we will be publishing for each task (i) annotations for a relatively small fine-tuning set, (ii) annotated development and test sets, and (iii) baseline models for easy reproducibility and comparisons. In this work, we present the details of data collection and annotation and the performance of the baseline models. We also perform sensitivity analysis of pipeline models' performance (speech recognizer + text model) to the speech recognition accuracy, using more than 20 state-of-the-art speech recognition models.

☆ Transformers Go for the LOLs: Generating (Humourous) Titles from Scientific Abstracts End-to-End

Yanran Chen, Steffen Eger

We consider the end-to-end abstract-to-title generation problem, exploring seven recent transformer based models (including ChatGPT) fine-tuned on more than 30k abstract-title pairs from NLP and machine learning venues. As an extension, we also consider the harder problem of generating humorous paper titles. For the latter, we compile the first large-scale humor annotated dataset for scientific papers in the NLP/ML domains, comprising almost 2.5k titles. We evaluate all models using human and automatic metrics. Our human evaluation suggests that our best end-to-end system performs similarly to human authors (but arguably slightly worse). Generating funny titles is more difficult, however, and our automatic systems clearly underperform relative to humans and often learn dataset artefacts of humor. Finally, ChatGPT, without any fine-tuning, performs on the level of our best fine-tuned system.

☆ Privacy-Preserving Domain Adaptation of Semantic Parsers

Fatemehsadat Mireshghallah, Richard Shin, Yu Su, Tatsunori Hashimoto, Jason Eisner

Task-oriented dialogue systems often assist users with personal or confidential matters. For this reason, the developers of such a system are generally prohibited from observing actual usage. So how can they know where the system is failing and needs more training data or new functionality? In this work, we study ways in which realistic user utterances can be generated synthetically, to help increase the linguistic and functional coverage of the system, without compromising the privacy of actual users. To this end, we propose a two-stage Differentially Private (DP) generation method which first generates latent semantic parses, and then generates utterances based on the parses. Our proposed approach improves MAUVE by 3.8$\times$ and parse tree node-type overlap by 1.4$\times$ relative to current approaches for private synthetic data generation, improving both on fluency and semantic coverage. We further validate our approach on a realistic domain adaptation task of adding new functionality from private user data to a semantic parser, and show gains of 1.3$\times$ on its accuracy with the new feature.

☆ CausalDialogue: Modeling Utterance-level Causality in Conversations

Yi-Lin Tuan, Alon Albalak, Wenda Xu, Michael Saxon, Connor Pryor, Lise Getoor, William Yang Wang

Despite their widespread adoption, neural conversation models have yet to exhibit natural chat capabilities with humans. In this research, we examine user utterances as causes and generated responses as effects, recognizing that changes in a cause should produce a different effect. To further explore this concept, we have compiled and expanded upon a new dataset called CausalDialogue through crowd-sourcing. This dataset includes multiple cause-effect pairs within a directed acyclic graph (DAG) structure. Our analysis reveals that traditional loss functions can struggle to effectively incorporate the DAG structure, leading us to propose a causality-enhanced method called Exponential Maximum Average Treatment Effect (ExMATE) to enhance the impact of causality at the utterance level in training neural conversation models. To evaluate the effectiveness of this approach, we have built a comprehensive benchmark using the CausalDialogue dataset leveraging large-scale pre-trained language models, and have assessed the results through both human and automatic evaluation metrics for coherence, diversity, and agility. Our findings show that current techniques are still unable to effectively address conversational DAGs, and that the ExMATE method can improve the diversity and agility of conventional loss functions while maintaining coherence.

☆ When Not to Trust Language Models: Investigating Effectiveness and Limitations of Parametric and Non-Parametric Memories

Alex Mallen, Akari Asai, Victor Zhong, Rajarshi Das, Hannaneh Hajishirzi, Daniel Khashabi

Despite their impressive performance on diverse tasks, large language models (LMs) still struggle with tasks requiring rich world knowledge, implying the limitations of relying solely on their parameters to encode a wealth of world knowledge. This paper aims to understand LMs' strengths and limitations in memorizing factual knowledge, by conducting large-scale knowledge probing experiments of 10 models and 4 augmentation methods on PopQA, our new open-domain QA dataset with 14k questions. We find that LMs struggle with less popular factual knowledge, and that scaling fails to appreciably improve memorization of factual knowledge in the tail. We then show that retrieval-augmented LMs largely outperform orders of magnitude larger LMs, while unassisted LMs remain competitive in questions about high-popularity entities. Based on those findings, we devise a simple, yet effective, method for powerful and efficient retrieval-augmented LMs, which retrieves non-parametric memories only when necessary. Experimental results show that this significantly improves models' performance while reducing the inference costs.

comment: Code and data available at https://github.com/AlexTMallen/adaptive-retrieval; work in progress

☆ Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questions

Harsh Trivedi, Niranjan Balasubramanian, Tushar Khot, Ashish Sabharwal

Recent work has shown that large language models are capable of generating natural language reasoning steps or Chains-of-Thoughts (CoT) to answer a multi-step question when prompted to do so. This is insufficient, however, when the necessary knowledge is not available or up-to-date within a model's parameters. A straightforward approach to address this is to retrieve text from an external knowledge source using the question as a query and prepend it as context to the model's input. This, however, is also insufficient for multi-step QA where \textit{what to retrieve} depends on \textit{what has already been derived}. To address this issue we propose IRCoT, a new approach that interleaves retrieval with CoT for multi-step QA, guiding the retrieval with CoT and in turn using retrieved results to improve CoT. Our experiments with GPT3 show substantial improvements in retrieval (up to 22 points) and downstream QA (up to 16 points) over the baselines on four datasets: HotpotQA, 2WikiMultihopQA, MuSiQue, and IIRC. Notably, our method also works well for much smaller models such as T5-Flan-large (0.7B) without any additional training.

☆ DePlot: One-shot visual language reasoning by plot-to-table translation

Fangyu Liu, Julian Martin Eisenschlos, Francesco Piccinno, Syrine Krichene, Chenxi Pang, Kenton Lee, Mandar Joshi, Wenhu Chen, Nigel Collier, Yasemin Altun

Visual language such as charts and plots is ubiquitous in the human world. Comprehending plots and charts requires strong reasoning skills. Prior state-of-the-art (SOTA) models require at least tens of thousands of training examples and their reasoning capabilities are still much limited, especially on complex human-written queries. This paper presents the first one-shot solution to visual language reasoning. We decompose the challenge of visual language reasoning into two steps: (1) plot-to-text translation, and (2) reasoning over the translated text. The key in this method is a modality conversion module, named as DePlot, which translates the image of a plot or chart to a linearized table. The output of DePlot can then be directly used to prompt a pretrained large language model (LLM), exploiting the few-shot reasoning capabilities of LLMs. To obtain DePlot, we standardize the plot-to-table task by establishing unified task formats and metrics, and train DePlot end-to-end on this task. DePlot can then be used off-the-shelf together with LLMs in a plug-and-play fashion. Compared with a SOTA model finetuned on more than >28k data points, DePlot+LLM with just one-shot prompting achieves a 24.0% improvement over finetuned SOTA on human-written queries from the task of chart QA.

☆ Can Current Task-oriented Dialogue Models Automate Real-world Scenarios in the Wild?

Sang-Woo Lee, Sungdong Kim, Donghyeon Ko, Donghoon Ham, Youngki Hong, Shin Ah Oh, Hyunhoon Jung, Wangkyo Jung, Kyunghyun Cho, Donghyun Kwak, Hyungsuk Noh, Woomyoung Park

Task-oriented dialogue (TOD) systems are mainly based on the slot-filling-based TOD (SF-TOD) framework, in which dialogues are broken down into smaller, controllable units (i.e., slots) to fulfill a specific task. A series of approaches based on this framework achieved remarkable success on various TOD benchmarks. However, we argue that the current TOD benchmarks are limited to surrogate real-world scenarios and that the current TOD models are still a long way from unraveling the scenarios. In this position paper, we first identify current status and limitations of SF-TOD systems. After that, we explore the WebTOD framework, the alternative direction for building a scalable TOD system when a web/mobile interface is available. In WebTOD, the dialogue system learns how to understand the web/mobile interface that the human agent interacts with, powered by a large-scale language model.

comment: Working in Progress

☆ Mini-Model Adaptation: Efficiently Extending Pretrained Models to New Languages via Aligned Shallow Training

Kelly Marchisio, Patrick Lewis, Yihong Chen, Mikel Artetxe

Prior work has shown that it is possible to expand pretrained Masked Language Models (MLMs) to new languages by learning a new set of embeddings, while keeping the transformer body frozen. Despite learning a small subset of parameters, this approach is not compute-efficient, as training the new embeddings requires a full forward and backward pass over the entire model. In this work, we propose mini-model adaptation, a compute-efficient alternative that builds a shallow mini-model from a fraction of a large model's parameters. New language-specific embeddings can then be efficiently trained over the mini-model, and plugged into the aligned large model for rapid cross-lingual transfer. We explore two approaches to learn mini-models: MiniJoint, which jointly pretrains the primary model and the mini-model using a single transformer with a secondary MLM head at a middle layer; and MiniPost, where we start from a regular pretrained model and build a mini-model by extracting and freezing a few layers and learning a small number of parameters on top. Experiments on XNLI, MLQA and PAWS-X show that mini-model adaptation matches the performance of the standard approach using up to 2.4x less compute.

☆ A Measure-Theoretic Characterization of Tight Language Models

Li Du, Lucas Torroba Hennigen, Tiago Pimentel, Clara Meister, Jason Eisner, Ryan Cotterell

Language modeling, a central task in natural language processing, involves estimating a probability distribution over strings. In most cases, the estimated distribution sums to 1 over all finite strings. However, in some pathological cases, probability mass can ``leak'' onto the set of infinite sequences. In order to characterize the notion of leakage more precisely, this paper offers a measure-theoretic treatment of language modeling. We prove that many popular language model families are in fact tight, meaning that they will not leak in this sense. We also generalize characterizations of tightness proposed in previous works.

☆ SimpleStyle: An Adaptable Style Transfer Approach

Elron Bandel, Yoav Katz, Noam Slonim, Liat Ein-Dor

Attribute Controlled Text Rewriting, also known as text style transfer, has received significant attention in the natural language generation community due to its crucial role in controllable natural language generation systems. In this work we present SimpleStyle a minimalist yet effective approach for attribute controlled text rewriting based on a simple mechanism composed of two ingredients. controlled denoising and output filtering. Despite the simplicity of our approach, which can be succinctly explained with just a few lines of code, it is competitive with previous state-of-the-art methods both in automatic and in human evaluations. Additionally, we demonstrate the practical effectiveness of our system, by applying it to real-world data from social networks. Additionally, we introduce a soft masking sampling technique that further improves the performance of the system. We also show that feeding the output of our system into a text-to-text student model can produce high-quality results without the need for additional filtering. Finally, we suggest that our method can solve the fundamental missing baseline absence that holding progress in the field by offering our protocol as a simple, adaptive and very strong baseline for works wish to make incremental advancements in the field of attribute controlled text rewriting.

☆ Precise Zero-Shot Dense Retrieval without Relevance Labels

Luyu Gao, Xueguang Ma, Jimmy Lin, Jamie Callan

While dense retrieval has been shown effective and efficient across tasks and languages, it remains difficult to create effective fully zero-shot dense retrieval systems when no relevance label is available. In this paper, we recognize the difficulty of zero-shot learning and encoding relevance. Instead, we propose to pivot through Hypothetical Document Embeddings~(HyDE). Given a query, HyDE first zero-shot instructs an instruction-following language model (e.g. InstructGPT) to generate a hypothetical document. The document captures relevance patterns but is unreal and may contain false details. Then, an unsupervised contrastively learned encoder~(e.g. Contriever) encodes the document into an embedding vector. This vector identifies a neighborhood in the corpus embedding space, where similar real documents are retrieved based on vector similarity. This second step ground the generated document to the actual corpus, with the encoder's dense bottleneck filtering out the incorrect details. Our experiments show that HyDE significantly outperforms the state-of-the-art unsupervised dense retriever Contriever and shows strong performance comparable to fine-tuned retrievers, across various tasks (e.g. web search, QA, fact verification) and languages~(e.g. sw, ko, ja).

☆ Execution-Based Evaluation for Open-Domain Code Generation

Zhiruo Wang, Shuyan Zhou, Daniel Fried, Graham Neubig

To extend the scope of coding queries to more realistic settings, we propose ODEX, the first open-domain execution-based natural language (NL) to code generation dataset. ODEX has 945 NL-Code pairs spanning 79 diverse libraries, along with 1,707 human-written test cases for execution. Our NL-Code pairs are harvested from StackOverflow forums to encourage natural and practical coding queries, which are then carefully rephrased to ensure intent clarity and prevent potential data memorization. Moreover, ODEX supports four natural languages as intents, in English, Spanish, Japanese, and Russian. ODEX unveils intriguing behavioral differences between top-performing Code LMs: Codex performs better on open-domain queries, yet CodeGen captures a better balance between open- and closed-domain. ODEX corroborates the merits of execution-based evaluation over metrics without execution but also unveils their complementary effects. Powerful models such as CodeGen-6B only achieve an 11.96 pass rate at top-1 prediction, suggesting plenty of headroom for improvement. We release ODEX to facilitate research into open-domain problems for the code generation community.

☆ ByGPT5: End-to-End Style-conditioned Poetry Generation with Token-free Language Models

Jonas Belouadi, Steffen Eger

State-of-the-art poetry generation systems are often complex. They either consist of task-specific model pipelines, incorporate prior knowledge in the form of manually created constraints or both. In contrast, end-to-end models would not suffer from the overhead of having to model prior knowledge and could learn the nuances of poetry from data alone, reducing the degree of human supervision required. In this work, we investigate end-to-end poetry generation conditioned on styles such as rhyme, meter, and alliteration. We identify and address lack of training data and mismatching tokenization algorithms as possible limitations of past attempts. In particular, we successfully pre-train and release ByGPT5, a new token-free decoder-only language model, and fine-tune it on a large custom corpus of English and German quatrains annotated with our styles. We show that ByGPT5 outperforms other models such as mT5, ByT5, GPT-2 and ChatGPT, while also being more parameter efficient and performing favorably compared to humans. In addition, we analyze its runtime performance and introspect the model's understanding of style conditions. We make our code, models, and datasets publicly available.

comment: Preprint

☆ Little Red Riding Hood Goes Around the Globe:Crosslingual Story Planning and Generation with Large Language Models

Evgeniia Razumovskaia, Joshua Maynez, Annie Louis, Mirella Lapata, Shashi Narayan

We consider the problem of automatically generating stories in multiple languages. Compared to prior work in monolingual story generation, crosslingual story generation allows for more universal research on story planning. We propose to use Prompting Large Language Models with Plans to study which plan is optimal for story generation. We consider 4 types of plans and systematically analyse how the outputs differ for different planning strategies. The study demonstrates that formulating the plans as question-answer pairs leads to more coherent generated stories while the plan gives more control to the story creators.

☆ BMX: Boosting Machine Translation Metrics with Explainability

Christoph Leiter, Hoa Nguyen, Steffen Eger

State-of-the-art machine translation evaluation metrics are based on black-box language models. Hence, recent works consider their explainability with the goals of better understandability for humans and better metric analysis, including failure cases. In contrast, we explicitly leverage explanations to boost the metrics' performance. In particular, we perceive explanations as word-level scores, which we convert, via power means, into sentence-level scores. We combine this sentence-level score with the original metric to obtain a better metric. Our extensive evaluation and analysis across 5 datasets, 5 metrics and 4 explainability techniques shows that some configurations reliably improve the original metrics' correlation with human judgment. On two held datasets for testing, we obtain improvements in 15/18 resp. 4/4 cases. The gains in Pearson correlation are up to 0.032 resp. 0.055. We make our code available.

comment: preprint

☆ Generic Temporal Reasoning with Differential Analysis and Explanation

Yu Feng, Ben Zhou, Haoyu Wang, Helen Jin, Dan Roth

Temporal reasoning is the task of predicting temporal relations of event pairs with corresponding contexts. While some temporal reasoning models perform reasonably well on in-domain benchmarks, we have little idea of the systems' generalizability due to existing datasets' limitations. In this work, we introduce a novel task named TODAY that bridges this gap with temporal differential analysis, which as the name suggests, evaluates if systems can correctly understand the effect of incremental changes. Specifically, TODAY makes slight context changes for given event pairs, and systems need to tell how this subtle contextual change will affect temporal relation distributions. To facilitate learning, TODAY also annotates human explanations. We show that existing models, including GPT-3, drop to random guessing on TODAY, suggesting that they heavily rely on spurious information rather than proper reasoning for temporal predictions. On the other hand, we show that TODAY's supervision style and explanation annotations can be used in joint learning and encourage models to use more appropriate signals during training and outperform across several benchmarks. TODAY can also be used to train models to solicit incidental supervision from noisy sources such as GPT-3 and moves farther towards generic temporal reasoning systems.

☆ Controllable Text Generation with Language Constraints

Howard Chen, Huihan Li, Danqi Chen, Karthik Narasimhan

We consider the task of text generation in language models with constraints specified in natural language. To this end, we first create a challenging benchmark Cognac that provides as input to the model a topic with example text, along with a constraint on text to be avoided. Unlike prior work, our benchmark contains knowledge-intensive constraints sourced from databases like Wordnet and Wikidata, which allows for straightforward evaluation while striking a balance between broad attribute-level and narrow lexical-level controls. We find that even state-of-the-art language models like GPT-3 fail often on this task, and propose a solution to leverage a language model's own internal knowledge to guide generation. Our method, called CognacGen, first queries the language model to generate guidance terms for a specified topic or constraint, and uses the guidance to modify the model's token generation probabilities. We propose three forms of guidance (binary verifier, top-k tokens, textual example), and employ prefix-tuning approaches to distill the guidance to tackle diverse natural language constraints. Through extensive empirical evaluations, we demonstrate that CognacGen can successfully generalize to unseen instructions and outperform competitive baselines in generating constraint conforming text.

☆ SODA: Million-scale Dialogue Distillation with Social Commonsense Contextualization

Hyunwoo Kim, Jack Hessel, Liwei Jiang, Ximing Lu, Youngjae Yu, Pei Zhou, Ronan Le Bras, Malihe Alikhani, Gunhee Kim, Maarten Sap, Yejin Choi

We present SODA: the first publicly available, million-scale high-quality social dialogue dataset. Using SODA, we train COSMO: a generalizable conversation agent outperforming previous best-performing agents on both in- and out-of-domain datasets. In contrast to most existing crowdsourced, small-scale dialogue corpora, we distill 1.5M socially-grounded dialogues from a pre-trained language model (InstructGPT; Ouyang et al., 2022). Dialogues are distilled by contextualizing social commonsense knowledge from a knowledge graph (Atomic10x; West et al., 2022). Human evaluation shows that dialogues in SODA are more consistent, specific, and (surprisingly) natural than prior human-authored datasets - e.g., DailyDialog (Li et al., 2017), BlendedSkillTalk (Smith et al., 2020). In addition, extensive evaluations show that COSMO is significantly more natural and consistent on unseen datasets than best-performing dialogue models - e.g., GODEL (Peng et al., 2022), BlenderBot (Roller et al., 2021), DialoGPT (Zhang et al., 2020). Furthermore, it is sometimes even preferred to the original human-written gold responses. We make our data, models, and code public.

comment: Dataset, models, and code can be found at https://hyunw.kim/sodaverse

☆ Go-tuning: Improving Zero-shot Learning Abilities of Smaller Language Models

Jingjing Xu, Qingxiu Dong, Hongyi Liu, Lei Li

With increasing scale, large language models demonstrate both quantitative improvement and new qualitative capabilities, especially as zero-shot learners, like GPT-3. However, these results rely heavily on delicate prompt design and large computation. In this work, we explore whether the strong zero-shot ability could be achieved at a smaller model scale without any external supervised data. To achieve this goal, we revisit masked language modeling and present a geometry-guided self-supervised learning method (Go-tuningfor short) by taking a small number of task-aware self-supervised data to update language models further. Experiments show that Go-tuning can enable T5-small (80M) competitive zero-shot results compared with large language models, such as T5-XL (3B). We also apply Go-tuning on multi-task settings and develop a multi-task model, mgo-T5 (250M). It can reach the average performance of OPT (175B) on 9 datasets.

☆ MULTI3NLU++: A Multilingual, Multi-Intent, Multi-Domain Dataset for Natural Language Understanding in Task-Oriented Dialogue

Nikita Moghe, Evgeniia Razumovskaia, Liane Guillou, Ivan Vulić, Anna Korhonen, Alexandra Birch

Task-oriented dialogue (TOD) systems have been applied in a range of domains to support human users to achieve specific goals. Systems are typically constructed for a single domain or language and do not generalise well beyond this. Their extension to other languages in particular is restricted by the lack of available training data for many of the world's languages. To support work on Natural Language Understanding (NLU) in TOD across multiple languages and domains simultaneously, we constructed MULTI3NLU++, a multilingual, multi-intent, multi-domain dataset. MULTI3NLU++ extends the English-only NLU++ dataset to include manual translations into a range of high, medium and low resource languages (Spanish, Marathi, Turkish and Amharic), in two domains (banking and hotels). MULTI3NLU++ inherits the multi-intent property of NLU++, where an utterance may be labelled with multiple intents, providing a more realistic representation of a user's goals and aligning with the more complex tasks that commercial systems aim to model. We use MULTI3NLU++ to benchmark state-of-the-art multilingual language models as well as Machine Translation and Question Answering systems for the NLU task of intent detection for TOD systems in the multilingual setting. The results demonstrate the challenging nature of the dataset, particularly in the low-resource language setting.

comment: Release of Dataset v1

☆ Is GPT-3 a Good Data Annotator?

Bosheng Ding, Chengwei Qin, Linlin Liu, Lidong Bing, Shafiq Joty, Boyang Li

GPT-3 (Generative Pre-trained Transformer 3) is a large-scale autoregressive language model developed by OpenAI, which has demonstrated impressive few-shot performance on a wide range of natural language processing (NLP) tasks. Hence, an intuitive application is to use it for data annotation. In this paper, we investigate whether GPT-3 can be used as a good data annotator for NLP tasks. Data annotation is the process of labeling data that could be used to train machine learning models. It is a crucial step in the development of NLP systems, as it allows the model to learn the relationship between the input data and the desired output. Given the impressive language capabilities of GPT-3, it is natural to wonder whether it can be used to effectively annotate data for NLP tasks. In this paper, we evaluate the performance of GPT-3 as a data annotator by comparing it with traditional data annotation methods and analyzing its output on a range of tasks. Through this analysis, we aim to provide insight into the potential of GPT-3 as a general-purpose data annotator in NLP.

☆ Socratic Pretraining: Question-Driven Pretraining for Controllable Summarization

Artidoro Pagnoni, Alexander R. Fabbri, Wojciech Kryściński, Chien-Sheng Wu

In long document controllable summarization, where labeled data is scarce, pretrained models struggle to adapt to the task and effectively respond to user queries. In this paper, we introduce Socratic pretraining, a question-driven, unsupervised pretraining objective specifically designed to improve controllability in summarization tasks. By training a model to generate and answer relevant questions in a given context, Socratic pretraining enables the model to more effectively adhere to user-provided queries and identify relevant content to be summarized. We demonstrate the effectiveness of this approach through extensive experimentation on two summarization domains, short stories and dialogue, and multiple control strategies: keywords, questions, and factoid QA pairs. Our pretraining method relies only on unlabeled documents and a question generation system and outperforms pre-finetuning approaches that use additional supervised data. Furthermore, our results show that Socratic pretraining cuts task-specific labeled data requirements in half, is more faithful to user-provided queries, and achieves state-of-the-art performance on QMSum and SQuALITY.

☆ Parameter-efficient Zero-shot Transfer for Cross-Language Dense Retrieval with Adapters

Eugene Yang, Suraj Nair, Dawn Lawrie, James Mayfield, Douglas W. Oard

A popular approach to creating a zero-shot cross-language retrieval model is to substitute a monolingual pretrained language model in the retrieval model with a multilingual pretrained language model such as Multilingual BERT. This multilingual model is fined-tuned to the retrieval task with monolingual data such as English MS MARCO using the same training recipe as the monolingual retrieval model used. However, such transferred models suffer from mismatches in the languages of the input text during training and inference. In this work, we propose transferring monolingual retrieval models using adapters, a parameter-efficient component for a transformer network. By adding adapters pretrained on language tasks for a specific language with task-specific adapters, prior work has shown that the adapter-enhanced models perform better than fine-tuning the entire model when transferring across languages in various NLP tasks. By constructing dense retrieval models with adapters, we show that models trained with monolingual data are more effective than fine-tuning the entire model when transferring to a Cross Language Information Retrieval (CLIR) setting. However, we found that the prior suggestion of replacing the language adapters to match the target language at inference time is suboptimal for dense retrieval models. We provide an in-depth analysis of this discrepancy between other cross-language NLP tasks and CLIR.

comment: 15 pages, 1 figure

☆ Perplexed by Quality: A Perplexity-based Method for Adult and Harmful Content Detection in Multilingual Heterogeneous Web Data

Tim Jansen, Yangling Tong, Victoria Zevallos, Pedro Ortiz Suarez

As demand for large corpora increases with the size of current state-of-the-art language models, using web data as the main part of the pre-training corpus for these models has become a ubiquitous practice. This, in turn, has introduced an important challenge for NLP practitioners, as they are now confronted with the task of developing highly optimized models and pipelines for pre-processing large quantities of textual data, which implies, effectively classifying and filtering multilingual, heterogeneous and noisy data, at web scale. One of the main components of this pre-processing step for the pre-training corpora of large language models, is the removal of adult and harmful content. In this paper we explore different methods for detecting adult and harmful of content in multilingual heterogeneous web data. We first show how traditional methods in harmful content detection, that seemingly perform quite well in small and specialized datasets quickly break down when confronted with heterogeneous noisy web data. We then resort to using a perplexity based approach but with a twist: Instead of using a so-called "clean" corpus to train a small language model and then use perplexity so select the documents with low perplexity, i.e., the documents that resemble this so-called "clean" corpus the most. We train solely with adult and harmful textual data, and then select the documents having a perplexity value above a given threshold. This approach will virtually cluster our documents into two distinct groups, which will greatly facilitate the choice of the threshold for the perplexity and will also allow us to obtain higher precision than with the traditional classification methods for detecting adult and harmful content.

comment: 14 pages, 2 figures

☆ Fine-Grained Distillation for Long Document Retrieval

Yucheng Zhou, Tao Shen, Xiubo Geng, Chongyang Tao, Guodong Long, Can Xu, Daxin Jiang

Long document retrieval aims to fetch query-relevant documents from a large-scale collection, where knowledge distillation has become de facto to improve a retriever by mimicking a heterogeneous yet powerful cross-encoder. However, in contrast to passages or sentences, retrieval on long documents suffers from the scope hypothesis that a long document may cover multiple topics. This maximizes their structure heterogeneity and poses a granular-mismatch issue, leading to an inferior distillation efficacy. In this work, we propose a new learning framework, fine-grained distillation (FGD), for long-document retrievers. While preserving the conventional dense retrieval paradigm, it first produces global-consistent representations crossing different fine granularity and then applies multi-granular aligned distillation merely during training. In experiments, we evaluate our framework on two long-document retrieval benchmarks, which show state-of-the-art performance.

comment: 13 pages, 5 figures, 5 tables

☆ Localising In-Domain Adaptation of Transformer-Based Biomedical Language Models

Tommaso Mario Buonocore, Claudio Crema, Enea Parimbelli, Alberto Redolfi, Riccardo Bellazzi

In the era of digital healthcare, the huge volumes of textual information generated every day in hospitals constitute an essential but underused asset that could be exploited with task-specific, fine-tuned biomedical language representation models, improving patient care and management. For such specialized domains, previous research has shown that fine-tuning models stemming from broad-coverage checkpoints can largely benefit additional training rounds over large-scale in-domain resources. However, these resources are often unreachable for less-resourced languages like Italian, preventing local medical institutions to employ in-domain adaptation. In order to reduce this gap, our work investigates two accessible approaches to derive biomedical language models in languages other than English, taking Italian as a concrete use-case: one based on neural machine translation of English resources, favoring quantity over quality; the other based on a high-grade, narrow-scoped corpus natively in Italian, thus preferring quality over quantity. Our study shows that data quantity is a harder constraint than data quality for biomedical adaptation, but the concatenation of high-quality data can improve model performance even when dealing with relatively size-limited corpora. The models published from our investigations have the potential to unlock important research opportunities for Italian hospitals and academia. Finally, the set of lessons learned from the study constitutes valuable insights towards a solution to build biomedical language models that are generalizable to other less-resourced languages and different domain settings.

☆ Reinforced Clarification Question Generation with Defeasibility Rewards for Disambiguating Social and Moral Situations

Valentina Pyatkin, Jena D. Hwang, Vivek Srikumar, Ximing Lu, Liwei Jiang, Yejin Choi, Chandra Bhagavatula

Context is vital for commonsense moral reasoning. "Lying to a friend" is wrong if it is meant to deceive them, but may be morally okay if it is intended to protect them. Such nuanced but salient contextual information can potentially flip the moral judgment of an action. Thus, we present ClarifyDelphi, an interactive system that elicits missing contexts of a moral situation by generating clarification questions such as "Why did you lie to your friend?". Our approach is inspired by the observation that questions whose potential answers lead to diverging moral judgments are the most informative. We learn to generate questions using Reinforcement Learning, by maximizing the divergence between moral judgements of hypothetical answers to a question. Human evaluation shows that our system generates more relevant, informative and defeasible questions compared to other question generation baselines. ClarifyDelphi assists informed moral reasoning processes by seeking additional morally consequential context to disambiguate social and moral situations.

comment: 9 pages + bibliography + appendix

☆ Geographic and Geopolitical Biases of Language Models

Fahim Faisal, Antonios Anastasopoulos

Pretrained language models (PLMs) often fail to fairly represent target users from certain world regions because of the under-representation of those regions in training datasets. With recent PLMs trained on enormous data sources, quantifying their potential biases is difficult, due to their black-box nature and the sheer scale of the data sources. In this work, we devise an approach to study the geographic bias (and knowledge) present in PLMs, proposing a Geographic-Representation Probing Framework adopting a self-conditioning method coupled with entity-country mappings. Our findings suggest PLMs' representations map surprisingly well to the physical world in terms of country-to-country associations, but this knowledge is unequally shared across languages. Last, we explain how large PLMs despite exhibiting notions of geographical proximity, over-amplify geopolitical favouritism at inference time.

☆ AnnoBERT: Effectively Representing Multiple Annotators' Label Choices to Improve Hate Speech Detection

Wenjie Yin, Vibhor Agarwal, Aiqi Jiang, Arkaitz Zubiaga, Nishanth Sastry

Supervised approaches generally rely on majority-based labels. However, it is hard to achieve high agreement among annotators in subjective tasks such as hate speech detection. Existing neural network models principally regard labels as categorical variables, while ignoring the semantic information in diverse label texts. In this paper, we propose AnnoBERT, a first-of-its-kind architecture integrating annotator characteristics and label text with a transformer-based model to detect hate speech, with unique representations based on each annotator's characteristics via Collaborative Topic Regression (CTR) and integrate label text to enrich textual representations. During training, the model associates annotators with their label choices given a piece of text; during evaluation, when label information is not available, the model predicts the aggregated label given by the participating annotators by utilising the learnt association. The proposed approach displayed an advantage in detecting hate speech, especially in the minority class and edge cases with annotator disagreement. Improvement in the overall performance is the largest when the dataset is more label-imbalanced, suggesting its practical value in identifying real-world hate speech, as the volume of hate speech in-the-wild is extremely small on social media, when compared with normal (non-hate) speech. Through ablation studies, we show the relative contributions of annotator embeddings and label text to the model performance, and tested a range of alternative annotator embeddings and label text combinations.

comment: accepted at ICWSM 2023

☆ Towards Reasoning in Large Language Models: A Survey

Jie Huang, Kevin Chen-Chuan Chang

Reasoning is a fundamental aspect of human intelligence that plays a crucial role in activities such as problem solving, decision making, and critical thinking. In recent years, large language models (LLMs) have made significant progress in natural language processing, and there is observation that these models may exhibit reasoning abilities when they are sufficiently large. However, it is not yet clear to what extent LLMs are capable of reasoning. This paper provides a comprehensive overview of the current state of knowledge on reasoning in LLMs, including techniques for improving and eliciting reasoning in these models, methods and benchmarks for evaluating reasoning abilities, findings and implications of previous research in this field, and suggestions on future directions. Our aim is to provide a detailed and up-to-date review of this topic and stimulate meaningful discussion and future work.

☆ Contrastive Learning Reduces Hallucination in Conversations AAAI2023

Weiwei Sun, Zhengliang Shi, Shen Gao, Pengjie Ren, Maarten de Rijke, Zhaochun Ren

Pre-trained language models (LMs) store knowledge in their parameters and can generate informative responses when used in conversational systems. However, LMs suffer from the problem of "hallucination:" they may generate plausible-looking statements that are irrelevant or factually incorrect. To address this problem, we propose a contrastive learning scheme, named MixCL. A novel mixed contrastive objective is proposed to explicitly optimize the implicit knowledge elicitation process of LMs, and thus reduce their hallucination in conversations. We also examine negative sampling strategies of retrieved hard negatives and model-generated negatives. We conduct experiments on Wizard-of-Wikipedia, a public, open-domain knowledge-grounded dialogue benchmark, and assess the effectiveness of MixCL. MixCL effectively reduces the hallucination of LMs in conversations and achieves the highest performance among LM-based dialogue agents in terms of relevancy and factuality. We show that MixCL achieves comparable performance to state-of-the-art KB-based approaches while enjoying notable advantages in terms of efficiency and scalability.

comment: Accepted by AAAI2023

☆ Needle in a Haystack: An Analysis of Finding Qualified Workers on MTurk for Summarization

Lining Zhang, João Sedoc, Simon Mille, Yufang Hou, Sebastian Gehrmann, Daniel Deutsch, Elizabeth Clark, Yixin Liu, Miruna Clinciu, Saad Mahamood, Khyathi Chandu

The acquisition of high-quality human annotations through crowdsourcing platforms like Amazon Mechanical Turk (MTurk) is more challenging than expected. The annotation quality might be affected by various aspects like annotation instructions, Human Intelligence Task (HIT) design, and wages paid to annotators, etc. To avoid potentially low-quality annotations which could mislead the evaluation of automatic summarization system outputs, we investigate the recruitment of high-quality MTurk workers via a three-step qualification pipeline. We show that we can successfully filter out bad workers before they carry out the evaluations and obtain high-quality annotations while optimizing the use of resources. This paper can serve as basis for the recruitment of qualified annotators in other challenging annotation tasks.

☆ Debiasing Stance Detection Models with Counterfactual Reasoning and Adversarial Bias Learning

Jianhua Yuan, Yanyan Zhao, Bing Qin

Stance detection models may tend to rely on dataset bias in the text part as a shortcut and thus fail to sufficiently learn the interaction between the targets and texts. Recent debiasing methods usually treated features learned by small models or big models at earlier steps as bias features and proposed to exclude the branch learning those bias features during inference. However, most of these methods fail to disentangle the ``good'' stance features and ``bad'' bias features in the text part. In this paper, we investigate how to mitigate dataset bias in stance detection. Motivated by causal effects, we leverage a novel counterfactual inference framework, which enables us to capture the dataset bias in the text part as the direct causal effect of the text on stances and reduce the dataset bias in the text part by subtracting the direct text effect from the total causal effect. We novelly model bias features as features that correlate with the stance labels but fail on intermediate stance reasoning subtasks and propose an adversarial bias learning module to model the bias more accurately. To verify whether our model could better model the interaction between texts and targets, we test our model on recently proposed test sets to evaluate the understanding of the task from various aspects. Experiments demonstrate that our proposed method (1) could better model the bias features, and (2) outperforms existing debiasing baselines on both the original dataset and most of the newly constructed test sets.

comment: Work in Progress

☆ TeSS: Zero-Shot Classification via Textual Similarity Comparison with Prompting using Sentence Encoder

Jimin Hong, Jungsoo Park, Daeyoung Kim, Seongjae Choi, Bokyung Son, Jaewook Kang

We introduce TeSS (Text Similarity Comparison using Sentence Encoder), a framework for zero-shot classification where the assigned label is determined by the embedding similarity between the input text and each candidate label prompt. We leverage representations from sentence encoders optimized to locate semantically similar samples closer to each other in embedding space during pre-training. The label prompt embeddings serve as prototypes of their corresponding class clusters. Furthermore, to compensate for the potentially poorly descriptive labels in their original format, we retrieve semantically similar sentences from external corpora and additionally use them with the original label prompt (TeSS-R). TeSS outperforms strong baselines on various closed-set and open-set classification datasets under zero-shot setting, with further gains when combined with label prompt diversification through retrieval. These results are robustly attained to verbalizer variations, an ancillary benefit of using a bi-encoder. Altogether, our method serves as a reliable baseline for zero-shot classification and a simple interface to assess the quality of sentence encoders.

comment: 9 pages, 3 figures

☆ To Adapt or to Annotate: Challenges and Interventions for Domain Adaptation in Open-Domain Question Answering

Dheeru Dua, Emma Strubell, Sameer Singh, Pat Verga

Recent advances in open-domain question answering (ODQA) have demonstrated impressive accuracy on standard Wikipedia style benchmarks. However, it is less clear how robust these models are and how well they perform when applied to real-world applications in drastically different domains. While there has been some work investigating how well ODQA models perform when tested for out-of-domain (OOD) generalization, these studies have been conducted only under conservative shifts in data distribution and typically focus on a single component (ie. retrieval) rather than an end-to-end system. In response, we propose a more realistic and challenging domain shift evaluation setting and, through extensive experiments, study end-to-end model performance. We find that not only do models fail to generalize, but high retrieval scores often still yield poor answer prediction accuracy. We then categorize different types of shifts and propose techniques that, when presented with a new dataset, predict if intervention methods are likely to be successful. Finally, using insights from this analysis, we propose and evaluate several intervention methods which improve end-to-end answer F1 score by up to 24 points.

☆ What Are You Token About? Dense Retrieval as Distributions Over the Vocabulary

Ori Ram, Liat Bezalel, Adi Zicher, Yonatan Belinkov, Jonathan Berant, Amir Globerson

Dual encoders are now the dominant architecture for dense retrieval. Yet, we have little understanding of how they represent text, and why this leads to good performance. In this work, we shed light on this question via distributions over the vocabulary. We propose to interpret the vector representations produced by dual encoders by projecting them into the model's vocabulary space. We show that the resulting distributions over vocabulary tokens are intuitive and contain rich semantic information. We find that this view can explain some of the failure cases of dense retrievers. For example, the inability of models to handle tail entities can be explained via a tendency of the token distributions to forget some of the tokens of those entities. We leverage this insight and propose a simple way to enrich query and passage representations with lexical information at inference time, and show that this significantly improves performance compared to the original model in out-of-domain settings.

☆ Careful Data Curation Stabilizes In-context Learning

Ting-Yun Chang, Robin Jia

In-context learning (ICL) enables large language models (LLMs) to perform new tasks by prompting them with a sequence of training examples. However, ICL is very sensitive to the choice of training examples: randomly sampling examples from a training set leads to high variance in performance. In this paper, we show that curating a carefully chosen subset of training data greatly stabilizes ICL performance. We propose two methods to choose training subsets, both of which score training examples individually and then select the highest-scoring ones. CondAcc scores a training example by its average ICL accuracy when combined with random training examples, while Datamodels learns a linear proxy model that estimates how the presence of each training example influences LLM accuracy. On average, CondAcc and Datamodels outperform sampling from the entire training set by 7.7% and 6.3%, respectively, across 5 tasks and two LLMs. Our analysis shows that stable subset examples are no more diverse than average, and are not outliers in terms of sequence length and perplexity.

☆ Self-adaptive In-context Learning

Zhiyong Wu, Yaoxiang Wang, Jiacheng Ye, Lingpeng Kong

Despite the surprising few-shot performance of in-context learning (ICL), it is still a common practice to randomly sample examples to serve as context. This paper advocates a new principle for ICL: self-adaptive in-context learning. The self-adaption mechanism is introduced to help each sample find an in-context example permutation (i.e., selection and ordering) that can derive the correct prediction, thus maximizing performance. To validate the effectiveness of self-adaptive ICL, we propose a general select-then-rank framework and instantiate it with new selection and ranking algorithms. Upon extensive evaluation on eight different NLP datasets, our self-adaptive ICL method achieves a 40% relative improvement over the common practice setting. Further analysis reveals the enormous potential of self-adaptive ICL that it might be able to close the gap between ICL and finetuning given more advanced algorithms. Our code is released to facilitate future research in this area: https://github.com/Shark-NLP/self-adaptive-ICL

☆ Receptive Field Alignment Enables Transformer Length Extrapolation

Ta-Chung Chi, Ting-Han Fan, Alexander I. Rudnicky

Length extrapolation is a desirable property that permits training a transformer language model on short sequences and retaining similar perplexities when the model is tested on substantially longer sequences. A relative positional embedding mechanism applied on the transformer self-attention matrix, ALiBi, demonstrates the length extrapolation property with the widest usage to date. In this paper, we show that ALiBi surprisingly does not utilize tokens further than the training sequence length, which can be explained by its implicit windowed attention effect that aligns the receptive field during training and testing stages. Inspired by ALiBi and the receptive filed alignment hypothesis, we propose another transformer positional embedding design named~\textbf{Sandwich} that uses longer than training sequence length information, and it is a greatly simplified formulation of the earliest proposed Sinusoidal positional embedding. Finally, we show that both ALiBi and Sandwich enable efficient inference thanks to their implicit windowed attention effect.

comment: Work In progress

☆ Does It Affect You? Social and Learning Implications of Using Cognitive-Affective State Recognition for Proactive Human-Robot Tutoring

Matthias Kraus, Diana Betancourt, Wolfgang Minker

Using robots in educational contexts has already shown to be beneficial for a student's learning and social behaviour. For levitating them to the next level of providing more effective and human-like tutoring, the ability to adapt to the user and to express proactivity is fundamental. By acting proactively, intelligent robotic tutors anticipate possible situations where problems for the student may arise and act in advance for preventing negative outcomes. Still, the decisions of when and how to behave proactively are open questions. Therefore, this paper deals with the investigation of how the student's cognitive-affective states can be used by a robotic tutor for triggering proactive tutoring dialogue. In doing so, it is aimed to improve the learning experience. For this reason, a concept learning task scenario was observed where a robotic assistant proactively helped when negative user states were detected. In a learning task, the user's states of frustration and confusion were deemed to have negative effects on the outcome of the task and were used to trigger proactive behaviour. In an empirical user study with 40 undergraduate and doctoral students, we studied whether the initiation of proactive behaviour after the detection of signs of confusion and frustration improves the student's concentration and trust in the agent. Additionally, we investigated which level of proactive dialogue is useful for promoting the student's concentration and trust. The results show that high proactive behaviour harms trust, especially when triggered during negative cognitive-affective states but contributes to keeping the student focused on the task when triggered in these states. Based on our study results, we further discuss future steps for improving the proactive assistance of robotic tutoring systems.

☆ CoCo: Coherence-Enhanced Machine-Generated Text Detection Under Data Limitation With Contrastive Learning

Xiaoming Liu, Zhaohan Zhang, Yichen Wang, Yu Lan, Chao Shen

Machine-Generated Text (MGT) detection, a task that discriminates MGT from Human-Written Text (HWT), plays a crucial role in preventing misuse of text generative models, which excel in mimicking human writing style recently. Latest proposed detectors usually take coarse text sequence as input and output some good results by fine-tune pretrained models with standard cross-entropy loss. However, these methods fail to consider the linguistic aspect of text (e.g., coherence) and sentence-level structures. Moreover, they lack the ability to handle the low-resource problem which could often happen in practice considering the enormous amount of textual data online. In this paper, we present a coherence-based contrastive learning model named CoCo to detect the possible MGT under low-resource scenario. Inspired by the distinctiveness and permanence properties of linguistic feature, we represent text as a coherence graph to capture its entity consistency, which is further encoded by the pretrained model and graph neural network. To tackle the challenges of data limitations, we employ a contrastive learning framework and propose an improved contrastive loss for making full use of hard negative samples in training stage. The experiment results on two public datasets prove our approach outperforms the state-of-art methods significantly.

☆ SeqDiffuSeq: Text Diffusion with Encoder-Decoder Transformers

Hongyi Yuan, Zheng Yuan, Chuanqi Tan, Fei Huang, Songfang Huang

Diffusion model, a new generative modelling paradigm, has achieved great success in image, audio, and video generation. However, considering the discrete categorical nature of text, it is not trivial to extend continuous diffusion models to natural language, and text diffusion models are less studied. Sequence-to-sequence text generation is one of the essential natural language processing topics. In this work, we apply diffusion models to approach sequence-to-sequence text generation, and explore whether the superiority generation performance of diffusion model can transfer to natural language domain. We propose SeqDiffuSeq, a text diffusion model for sequence-to-sequence generation. SeqDiffuSeq uses an encoder-decoder Transformers architecture to model denoising function. In order to improve generation quality, SeqDiffuSeq combines the self-conditioning technique and a newly proposed adaptive noise schedule technique. The adaptive noise schedule has the difficulty of denoising evenly distributed across time steps, and considers exclusive noise schedules for tokens at different positional order. Experiment results illustrate the good performance on sequence-to-sequence generation in terms of text quality and inference time.

comment: Working in progress; 9 pages, 2 figures

☆ HINT: Hypernetwork Instruction Tuning for Efficient Zero-Shot Generalisation

Hamish Ivison, Akshita Bhagia, Yizhong Wang, Hannaneh Hajishirzi, Matthew Peters

Recent NLP models have the great ability to generalise `zero-shot' to new tasks using only an instruction as guidance. However, these approaches usually repeat their instructions with every input, requiring costly reprocessing of lengthy instructions for every inference example. To alleviate this, we introduce Hypernetworks for INstruction Tuning (HINT), which convert task instructions and examples using a pretrained text encoder into parameter-efficient modules inserted into an underlying model, eliminating the need to include instructions in the model input. Compared to prior approaches that concatenate instructions with every input instance, we find that HINT models are significantly more compute-efficient and consistently outperform these approaches for a given inference budget.

☆ Beyond Triplet: Leveraging the Most Data for Multimodal Machine Translation

Yaoming Zhu, Zewei Sun, Shanbo Cheng, Yuyang Huang, Liwei Wu, Mingxuan Wang

Multimodal machine translation (MMT) aims to improve translation quality by incorporating information from other modalities, such as vision. Previous MMT systems mainly focus on better access and use of visual information and tend to validate their methods on image-related datasets. These studies face two challenges. First, they can only utilize triple data (bilingual texts with images), which is scarce; second, current benchmarks are relatively restricted and do not correspond to realistic scenarios. Therefore, this paper correspondingly establishes new methods and new datasets for MMT. First, we propose a framework 2/3-Triplet with two new approaches to enhance MMT by utilizing large-scale non-triple data: monolingual image-text data and parallel text-only data. Second, we construct an English-Chinese {e}-commercial {m}ulti{m}odal {t}ranslation dataset (including training and testing), named EMMT, where its test set is carefully selected as some words are ambiguous and shall be translated mistakenly without the help of images. Experiments show that our method is more suitable for real-world scenarios and can significantly improve translation performance by using more non-triple data. In addition, our model also rivals various SOTA models in conventional multimodal translation benchmarks.

comment: 8 pages

☆ Extrinsic Evaluation of Machine Translation Metrics

Nikita Moghe, Tom Sherborne, Mark Steedman, Alexandra Birch

Automatic machine translation (MT) metrics are widely used to distinguish the translation qualities of machine translation systems across relatively large test sets (system-level evaluation). However, it is unclear if automatic metrics are reliable at distinguishing good translations from bad translations at the sentence level (segment-level evaluation). In this paper, we investigate how useful MT metrics are at detecting the success of a machine translation component when placed in a larger platform with a downstream task. We evaluate the segment-level performance of the most widely used MT metrics (chrF, COMET, BERTScore, etc.) on three downstream cross-lingual tasks (dialogue state tracking, question answering, and semantic parsing). For each task, we only have access to a monolingual task-specific model. We calculate the correlation between the metric's ability to predict a good/bad translation with the success/failure on the final task for the Translate-Test setup. Our experiments demonstrate that all metrics exhibit negligible correlation with the extrinsic evaluation of the downstream outcomes. We also find that the scores provided by neural metrics are not interpretable mostly because of undefined ranges. Our analysis suggests that future MT metrics be designed to produce error labels rather than scores to facilitate extrinsic evaluation.

comment: Work in Progress

☆ ReCode: Robustness Evaluation of Code Generation Models

Shiqi Wang, Zheng Li, Haifeng Qian, Chenghao Yang, Zijian Wang, Mingyue Shang, Varun Kumar, Samson Tan, Baishakhi Ray, Parminder Bhatia, Ramesh Nallapati, Murali Krishna Ramanathan, Dan Roth, Bing Xiang

Code generation models have achieved impressive performance. However, they tend to be brittle as slight edits to a prompt could lead to very different generations; these robustness properties, critical for user experience when deployed in real-life applications, are not well understood. Most existing works on robustness in text or code tasks have focused on classification, while robustness in generation tasks is an uncharted area and to date there is no comprehensive benchmark for robustness in code generation. In this paper, we propose ReCode, a comprehensive robustness evaluation benchmark for code generation models. We customize over 30 transformations specifically for code on docstrings, function and variable names, code syntax, and code format. They are carefully designed to be natural in real-life coding practice, preserve the original semantic meaning, and thus provide multifaceted assessments of a model's robustness performance. With human annotators, we verified that over 90% of the perturbed prompts do not alter the semantic meaning of the original prompt. In addition, we define robustness metrics for code generation models considering the worst-case behavior under each type of perturbation, taking advantage of the fact that executing the generated code can serve as objective evaluation. We demonstrate ReCode on SOTA models using HumanEval, MBPP, as well as function completion tasks derived from them. Interesting observations include: better robustness for CodeGen over InCoder and GPT-J; models are most sensitive to syntax perturbations; more challenging robustness evaluation on MBPP over HumanEval.

comment: Code and data available at https://github.com/amazon-science/recode

☆ In and Out-of-Domain Text Adversarial Robustness via Label Smoothing

Yahan Yang, Soham Dan, Dan Roth, Insup Lee

Recently it has been shown that state-of-the-art NLP models are vulnerable to adversarial attacks, where the predictions of a model can be drastically altered by slight modifications to the input (such as synonym substitutions). While several defense techniques have been proposed, and adapted, to the discrete nature of text adversarial attacks, the benefits of general-purpose regularization methods such as label smoothing for language models, have not been studied. In this paper, we study the adversarial robustness provided by various label smoothing strategies in foundational models for diverse NLP tasks in both in-domain and out-of-domain settings. Our experiments show that label smoothing significantly improves adversarial robustness in pre-trained models like BERT, against various popular attacks. We also analyze the relationship between prediction confidence and robustness, showing that label smoothing reduces over-confident errors on adversarial examples.

comment: Preprint. Under Submission

☆ Original or Translated? On the Use of Parallel Data for Translation Quality Estimation

Baopu Qiu, Liang Ding, Di Wu, Lin Shang, Yibing Zhan, Dacheng Tao

Machine Translation Quality Estimation (QE) is the task of evaluating translation output in the absence of human-written references. Due to the scarcity of human-labeled QE data, previous works attempted to utilize the abundant unlabeled parallel corpora to produce additional training data with pseudo labels. In this paper, we demonstrate a significant gap between parallel data and real QE data: for QE data, it is strictly guaranteed that the source side is original texts and the target side is translated (namely translationese). However, for parallel data, it is indiscriminate and the translationese may occur on either source or target side. We compare the impact of parallel data with different translation directions in QE data augmentation, and find that using the source-original part of parallel corpus consistently outperforms its target-original counterpart. Moreover, since the WMT corpus lacks direction information for each parallel sentence, we train a classifier to distinguish source- and target-original bitext, and carry out an analysis of their difference in both style and domain. Together, these findings suggest using source-original parallel data for QE data augmentation, which brings a relative improvement of up to 4.0% and 6.4% compared to undifferentiated data on sentence- and word-level QE tasks respectively.

comment: work in progress

☆ Diff-Glat: Diffusion Glancing Transformer for Parallel Sequence to Sequence Learning

Lihua Qian, Mingxuan Wang, Yang Liu, Hao Zhou

For sequence generation, both autoregressive models and non-autoregressive models have been developed in recent years. Autoregressive models can achieve high generation quality, but the sequential decoding scheme causes slow decoding speed. Non-autoregressive models accelerate the inference speed with parallel decoding, while their generation quality still needs to be improved due to the difficulty of modeling multi-modalities in data. To address the multi-modality issue, we propose Diff-Glat, a non-autoregressive model featured with a modality diffusion process and residual glancing training. The modality diffusion process decomposes the modalities and reduces the modalities to learn for each transition. And the residual glancing sampling further smooths the modality learning procedures. Experiments demonstrate that, without using knowledge distillation data, Diff-Glat can achieve superior performance in both decoding efficiency and accuracy compared with the autoregressive Transformer.

comment: Work in Progress

☆ Pre-trained Language Models for Keyphrase Generation: A Thorough Empirical Study

Di Wu, Wasi Uddin Ahmad, Kai-Wei Chang

Neural models that do not rely on pre-training have excelled in the keyphrase generation task with large annotated datasets. Meanwhile, new approaches have incorporated pre-trained language models (PLMs) for their data efficiency. However, there lacks a systematic study of how the two types of approaches compare and how different design choices can affect the performance of PLM-based models. To fill in this knowledge gap and facilitate a more informed use of PLMs for keyphrase extraction and keyphrase generation, we present an in-depth empirical study. Formulating keyphrase extraction as sequence labeling and keyphrase generation as sequence-to-sequence generation, we perform extensive experiments in three domains. After showing that PLMs have competitive high-resource performance and state-of-the-art low-resource performance, we investigate important design choices including in-domain PLMs, PLMs with different pre-training objectives, using PLMs with a parameter budget, and different formulations for present keyphrases. Further results show that (1) in-domain BERT-like PLMs can be used to build strong and data-efficient keyphrase generation models; (2) with a fixed parameter budget, prioritizing model depth over width and allocating more layers in the encoder leads to better encoder-decoder models; and (3) introducing four in-domain PLMs, we achieve a competitive performance in the news domain and the state-of-the-art performance in the scientific domain.

☆ GanLM: Encoder-Decoder Pre-training with an Auxiliary Discriminator

Jian Yang, Shuming Ma, Li Dong, Shaohan Huang, Haoyang Huang, Yuwei Yin, Dongdong Zhang, Liqun Yang, Zhoujun Li, Furu Wei

Pre-trained models have achieved remarkable success in natural language processing (NLP). However, existing pre-training methods underutilize the benefits of language understanding for generation. Inspired by the idea of Generative Adversarial Networks (GANs), we propose a GAN-style model for encoder-decoder pre-training by introducing an auxiliary discriminator, unifying the ability of language understanding and generation in a single model. Our model, named as GanLM, is trained with two pre-training objectives: replaced token detection and replaced token denoising. Specifically, given masked source sentences, the generator outputs the target distribution and the discriminator predicts whether the target sampled tokens from distribution are incorrect. The target sentence is replaced with misclassified tokens to construct noisy previous context, which is used to generate the gold sentence. In general, both tasks improve the ability of language understanding and generation by selectively using the denoising data. Extensive experiments in language generation benchmarks show that GanLM with the powerful language understanding capability outperforms various strong pre-trained language models (PLMs) and achieves state-of-the-art performance.

☆ EIT: Enhanced Interactive Transformer

Tong Zheng, Bei Li, Huiwen Bao, Tong Xiao, Jingbo Zhu

In this paper, we propose a novel architecture, the Enhanced Interactive Transformer (EIT), to address the issue of head degradation in self-attention mechanisms. Our approach replaces the traditional multi-head self-attention mechanism with the Enhanced Multi-Head Attention (EMHA) mechanism, which relaxes the one-to-one mapping constraint among queries and keys, allowing each query to attend to multiple keys. Furthermore, we introduce two interaction models, Inner-Subspace Interaction and Cross-Subspace Interaction, to fully utilize the many-to-many mapping capabilities of EMHA. Extensive experiments on a wide range of tasks (e.g. machine translation, abstractive summarization, grammar correction, language modelling and brain disease automatic diagnosis) show its superiority with a very modest increase in model size.

comment: 25 pages, 21 figures

☆ Adam: Dense Retrieval Distillation with Adaptive Dark Examples

Chang Liu, Chongyang Tao, Xiubo Geng, Tao Shen, Dongyan Zhao, Can Xu, Binxing Jiao, Daxin Jiang

To improve the performance of the dual-encoder retriever, one effective approach is knowledge distillation from the cross-encoder ranker. Existing works construct the candidate passages following the supervised learning setting where a query is paired with a positive passage and a batch of negatives. However, through empirical observation, we find that even the hard negatives from advanced methods are still too trivial for the teacher to distinguish, preventing the teacher from transferring abundant dark knowledge to the student through its soft label. To alleviate this issue, we propose ADAM, a knowledge distillation framework that can better transfer the dark knowledge held in the teacher with Adaptive Dark exAMples. Different from previous works that only rely on one positive and hard negatives as candidate passages, we create dark examples that all have moderate relevance to the query through mixing-up and masking in discrete space. Furthermore, as the quality of knowledge held in different training instances varies as measured by the teacher's confidence score, we propose a self-paced distillation strategy that adaptively concentrates on a subset of high-quality instances to conduct our dark-example-based knowledge distillation to help the student learn better. We conduct experiments on two widely-used benchmarks and verify the effectiveness of our method.

comment: 9 pages, 2 figures

☆ Emotion Selectable End-to-End Text-based Speech Editing

Tao Wang, Jiangyan Yi, Ruibo Fu, Jianhua Tao, Zhengqi Wen, Chu Yuan Zhang

Text-based speech editing allows users to edit speech by intuitively cutting, copying, and pasting text to speed up the process of editing speech. In the previous work, CampNet (context-aware mask prediction network) is proposed to realize text-based speech editing, significantly improving the quality of edited speech. This paper aims at a new task: adding emotional effect to the editing speech during the text-based speech editing to make the generated speech more expressive. To achieve this task, we propose Emo-CampNet (emotion CampNet), which can provide the option of emotional attributes for the generated speech in text-based speech editing and has the one-shot ability to edit unseen speakers' speech. Firstly, we propose an end-to-end emotion-selectable text-based speech editing model. The key idea of the model is to control the emotion of generated speech by introducing additional emotion attributes based on the context-aware mask prediction network. Secondly, to prevent the emotion of the generated speech from being interfered by the emotional components in the original speech, a neutral content generator is proposed to remove the emotion from the original speech, which is optimized by the generative adversarial framework. Thirdly, two data augmentation methods are proposed to enrich the emotional and pronunciation information in the training set, which can enable the model to edit the unseen speaker's speech. The experimental results that 1) Emo-CampNet can effectively control the emotion of the generated speech in the process of text-based speech editing; And can edit unseen speakers' speech. 2) Detailed ablation experiments further prove the effectiveness of emotional selectivity and data augmentation methods. The demo page is available at https://hairuo55.github.io/Emo-CampNet/

comment: Under review, 12 pages, 11 figures, demo page is available at https://hairuo55.github.io/Emo-CampNet/

☆ Pay Attention to Your Tone: Introducing a New Dataset for Polite Language Rewrite

Xun Wang, Tao Ge, Allen Mao, Yuki Li, Furu Wei, Si-Qing Chen

We introduce \textsc{PoliteRewrite} -- a dataset for polite language rewrite which is a novel sentence rewrite task. Compared with previous text style transfer tasks that can be mostly addressed by slight token- or phrase-level edits, polite language rewrite requires deep understanding and extensive sentence-level edits over an offensive and impolite sentence to deliver the same message euphemistically and politely, which is more challenging -- not only for NLP models but also for human annotators to rewrite with effort. To alleviate the human effort for efficient annotation, we first propose a novel annotation paradigm by a collaboration of human annotators and GPT-3.5 to annotate \textsc{PoliteRewrite}. The released dataset has 10K polite sentence rewrites annotated collaboratively by GPT-3.5 and human, which can be used as gold standard for training, validation and test; and 100K high-quality polite sentence rewrites by GPT-3.5 without human review. We wish this work (The dataset (10K+100K) will be released soon) could contribute to the research on more challenging sentence rewrite, and provoke more thought in future on resource annotation paradigm with the help of the large-scaled pretrained models.

☆ Do I have the Knowledge to Answer? Investigating Answerability of Knowledge Base Questions

Mayur Patidar, Avinash Singh, Prayushi Faldu, Lovekesh Vig, Indrajit Bhattacharya, Mausam

When answering natural language questions over knowledge bases (KBs), incompleteness in the KB can naturally lead to many questions being unanswerable. While answerability has been explored in other QA settings, it has not been studied for QA over knowledge bases (KBQA). We first identify various forms of KB incompleteness that can result in a question being unanswerable. We then propose GrailQAbility, a new benchmark dataset, which systematically modifies GrailQA (a popular KBQA dataset) to represent all these incompleteness issues. Testing two state-of-the-art KBQA models (trained on original GrailQA as well as our GrailQAbility), we find that both models struggle to detect unanswerable questions, or sometimes detect them for the wrong reasons. Consequently, both models suffer significant loss in performance, underscoring the need for further research in making KBQA systems robust to unanswerability.

☆ IndicMT Eval: A Dataset to Meta-Evaluate Machine Translation metrics for Indian Languages

Ananya B. Sai, Vignesh Nagarajan, Tanay Dixit, Raj Dabre, Anoop Kunchukuttan, Pratyush Kumar, Mitesh M. Khapra

The rapid growth of machine translation (MT) systems has necessitated comprehensive studies to meta-evaluate evaluation metrics being used, which enables a better selection of metrics that best reflect MT quality. Unfortunately, most of the research focuses on high-resource languages, mainly English, the observations for which may not always apply to other languages. Indian languages, having over a billion speakers, are linguistically different from English, and to date, there has not been a systematic study of evaluating MT systems from English into Indian languages. In this paper, we fill this gap by creating an MQM dataset consisting of 7000 fine-grained annotations, spanning 5 Indian languages and 7 MT systems, and use it to establish correlations between annotator scores and scores obtained using existing automatic metrics. Our results show that pre-trained metrics, such as COMET, have the highest correlations with annotator scores. Additionally, we find that the metrics do not adequately capture fluency-based errors in Indian languages, and there is a need to develop metrics focused on Indian languages. We hope that our dataset and analysis will help promote further research in this area.

☆ Toward Human-Like Evaluation for Natural Language Generation with Error Analysis

Qingyu Lu, Liang Ding, Liping Xie, Kanjian Zhang, Derek F. Wong, Dacheng Tao

The state-of-the-art language model-based automatic metrics, e.g. BARTScore, benefiting from large-scale contextualized pre-training, have been successfully used in a wide range of natural language generation (NLG) tasks, including machine translation, text summarization, and data-to-text. Recent studies show that considering both major errors (e.g. mistranslated tokens) and minor errors (e.g. imperfections in fluency) can produce high-quality human judgments. This inspires us to approach the final goal of the evaluation metrics (human-like evaluations) by automatic error analysis. To this end, we augment BARTScore by incorporating the human-like error analysis strategies, namely BARTScore++, where the final score consists of both the evaluations of major errors and minor errors. Experimental results show that BARTScore++ can consistently improve the performance of vanilla BARTScore and outperform existing top-scoring metrics in 20 out of 25 test settings. We hope our technique can also be extended to other pre-trained model-based metrics. We will release our code and scripts to facilitate the community.

comment: work in progress

☆ On the Role of Parallel Data in Cross-lingual Transfer Learning

Machel Reid, Mikel Artetxe

While prior work has established that the use of parallel data is conducive for cross-lingual learning, it is unclear if the improvements come from the data itself, or if it is the modeling of parallel interactions that matters. Exploring this, we examine the usage of unsupervised machine translation to generate synthetic parallel data, and compare it to supervised machine translation and gold parallel data. We find that even model generated parallel data can be useful for downstream tasks, in both a general setting (continued pretraining) as well as the task-specific setting (translate-train), although our best results are still obtained using real parallel data. Our findings suggest that existing multilingual models do not exploit the full potential of monolingual data, and prompt the community to reconsider the traditional categorization of cross-lingual learning approaches.

comment: Preprint

☆ Document-level Relation Extraction with Relation Correlations

Ridong Han, Tao Peng, Benyou Wang, Lu Liu, Xiang Wan

Document-level relation extraction faces two overlooked challenges: long-tail problem and multi-label problem. Previous work focuses mainly on obtaining better contextual representations for entity pairs, hardly address the above challenges. In this paper, we analyze the co-occurrence correlation of relations, and introduce it into DocRE task for the first time. We argue that the correlations can not only transfer knowledge between data-rich relations and data-scarce ones to assist in the training of tailed relations, but also reflect semantic distance guiding the classifier to identify semantically close relations for multi-label entity pairs. Specifically, we use relation embedding as a medium, and propose two co-occurrence prediction sub-tasks from both coarse- and fine-grained perspectives to capture relation correlations. Finally, the learned correlation-aware embeddings are used to guide the extraction of relational facts. Substantial experiments on two popular DocRE datasets are conducted, and our method achieves superior results compared to baselines. Insightful analysis also demonstrates the potential of relation correlations to address the above challenges.

comment: 13 pages

☆ Naamapadam: A Large-Scale Named Entity Annotated Data for Indic Languages

Arnav Mhaske, Harshit Kedia, Sumanth Doddapaneni, Mitesh M. Khapra, Pratyush Kumar, Rudra Murthy V, Anoop Kunchukuttan

We present, Naamapadam, the largest publicly available Named Entity Recognition (NER) dataset for the 11 major Indian languages from two language families. In each language, it contains more than 400k sentences annotated with a total of at least 100k entities from three standard entity categories (Person, Location and Organization) for 9 out of the 11 languages. The training dataset has been automatically created from the Samanantar parallel corpus by projecting automatically tagged entities from an English sentence to the corresponding Indian language sentence. We also create manually annotated testsets for 8 languages containing approximately 1000 sentences per language. We demonstrate the utility of the obtained dataset on existing testsets and the Naamapadam-test data for 8 Indic languages. We also release IndicNER, a multilingual mBERT model fine-tuned on the Naamapadam training set. IndicNER achieves the best F1 on the Naamapadam-test set compared to an mBERT model fine-tuned on existing datasets. IndicNER achieves an F1 score of more than 80 for 7 out of 11 Indic languages. The dataset and models are available under open-source licenses at https://ai4bharat.iitm.ac.in/naamapadam.

comment: 14 pages, 5 figures, Work in Progress

☆ Human-Guided Fair Classification for Natural Language Processing

Florian E. Dorner, Momchil Peychev, Nikola Konstantinov, Naman Goel, Elliott Ash, Martin Vechev

Text classifiers have promising applications in high-stake tasks such as resume screening and content moderation. These classifiers must be fair and avoid discriminatory decisions by being invariant to perturbations of sensitive attributes such as gender or ethnicity. However, there is a gap between human intuition about these perturbations and the formal similarity specifications capturing them. While existing research has started to address this gap, current methods are based on hardcoded word replacements, resulting in specifications with limited expressivity or ones that fail to fully align with human intuition (e.g., in cases of asymmetric counterfactuals). This work proposes novel methods for bridging this gap by discovering expressive and intuitive individual fairness specifications. We show how to leverage unsupervised style transfer and GPT-3's zero-shot capabilities to automatically generate expressive candidate pairs of semantically similar sentences that differ along sensitive attributes. We then validate the generated pairs via an extensive crowdsourcing study, which confirms that a lot of these pairs align with human intuition about fairness in the context of toxicity classification. Finally, we show how limited amounts of human feedback can be leveraged to learn a similarity specification that can be used to train downstream fairness-aware models.

comment: 31 pages, 1 figure

☆ Quirk or Palmer: A Comparative Study of Modal Verb Frameworks with Annotated Datasets

Risako Owan, Maria Gini, Dongyeop Kang

Modal verbs, such as "can", "may", and "must", are commonly used in daily communication to convey the speaker's perspective related to the likelihood and/or mode of the proposition. They can differ greatly in meaning depending on how they're used and the context of a sentence (e.g. "They 'must' help each other out." vs. "They 'must' have helped each other out.") Despite their practical importance in natural language understanding, linguists have yet to agree on a single, prominent framework for the categorization of modal verb senses. This lack of agreement stems from high degrees of flexibility and polysemy from the modal verbs, making it more difficult for researchers to incorporate insights from this family of words into their work. This work presents Moverb dataset, which consists of 27,240 annotations of modal verb senses over 4,540 utterances containing one or more sentences from social conversations. Each utterance is annotated by three annotators using two different theoretical frameworks (i.e., Quirk and Palmer) of modal verb senses. We observe that both frameworks have similar inter-annotator agreements, despite having different numbers of sense types (8 for Quirk and 3 for Palmer). With the RoBERTa-based classifiers fine-tuned on \dataset, we achieve F1 scores of 82.2 and 78.3 on Quirk and Palmer, respectively, showing that modal verb sense disambiguation is not a trivial task. Our dataset will be publicly available with our final version.

☆ Tackling Ambiguity with Images: Improved Multimodal Machine Translation and Contrastive Evaluation

Matthieu Futeral, Cordelia Schmid, Ivan Laptev, Benoît Sagot, Rachel Bawden

One of the major challenges of machine translation (MT) is ambiguity, which can in some cases be resolved by accompanying context such as an image. However, recent work in multimodal MT (MMT) has shown that obtaining improvements from images is challenging, limited not only by the difficulty of building effective cross-modal representations but also by the lack of specific evaluation and training data. We present a new MMT approach based on a strong text-only MT model, which uses neural adapters and a novel guided self-attention mechanism and which is jointly trained on both visual masking and MMT. We also release CoMMuTE, a Contrastive Multilingual Multimodal Translation Evaluation dataset, composed of ambiguous sentences and their possible translations, accompanied by disambiguating images corresponding to each translation. Our approach obtains competitive results over strong text-only models on standard English-to-French benchmarks and outperforms these baselines and state-of-the-art MMT systems with a large margin on our contrastive test set.

☆ True Detective: A Challenging Benchmark for Deep Abductive Reasoning \\in Foundation Models

Maksym Del, Mark Fishel

Large language models (LLMs) have demonstrated strong performance in zero-shot reasoning tasks, including abductive reasoning. This is reflected in their ability to perform well on current benchmarks in this area. However, to truly test the limits of LLMs in abductive reasoning, a more challenging benchmark is needed. In this paper, we present such a benchmark, consisting of 191 long-form mystery stories, each approximately 1200 words in length and presented in the form of detective puzzles. Each puzzle includes a multiple-choice question for evaluation sourced from the "5 Minute Mystery" platform. Our results show that state-of-the-art GPT models perform significantly worse than human solvers on this benchmark, with an accuracy of 28\% compared to 47\% for humans. This indicates that there is still a significant gap in the abductive reasoning abilities of LLMs and highlights the need for further research in this area. Our work provides a challenging benchmark for future studies on reasoning in language models and contributes to a better understanding of the limits of LLMs' abilities.

comment: 4 pages, preprint

☆ Toward a Unified Framework for Unsupervised Complex Tabular Reasoning ICDE 2023

Zhenyu Li, Xiuxing Li, Zhichao Duan, Bowen Dong, Ning Liu, Jianyong Wang

Structured tabular data exist across nearly all fields. Reasoning task over these data aims to answer questions or determine the truthiness of hypothesis sentences by understanding the semantic meaning of a table. While previous works have devoted significant efforts to the tabular reasoning task, they always assume there are sufficient labeled data. However, constructing reasoning samples over tables (and related text) is labor-intensive, especially when the reasoning process is complex. When labeled data is insufficient, the performance of models will suffer an unendurable decline. In this paper, we propose a unified framework for unsupervised complex tabular reasoning (UCTR), which generates sufficient and diverse synthetic data with complex logic for tabular reasoning tasks, assuming no human-annotated data at all. We first utilize a random sampling strategy to collect diverse programs of different types and execute them on tables based on a "Program-Executor" module. To bridge the gap between the programs and natural language sentences, we design a powerful "NL-Generator" module to generate natural language sentences with complex logic from these programs. Since a table often occurs with its surrounding texts, we further propose novel "Table-to-Text" and "Text-to-Table" operators to handle joint table-text reasoning scenarios. This way, we can adequately exploit the unlabeled table resources to obtain a well-performed reasoning model under an unsupervised setting. Our experiments cover different tasks (question answering and fact verification) and different domains (general and specific), showing that our unsupervised methods can achieve at most 93% performance compared to supervised models. We also find that it can substantially boost the supervised performance in low-resourced domains as a data augmentation technique. Our code is available at https://github.com/leezythu/UCTR.

comment: Accepted by ICDE 2023. Preprint Version

☆ Hybrid Rule-Neural Coreference Resolution System based on Actor-Critic Learning

Yu Wang, Hongxia Jin

A coreference resolution system is to cluster all mentions that refer to the same entity in a given context. All coreference resolution systems need to tackle two main tasks: one task is to detect all of the potential mentions, and the other is to learn the linking of an antecedent for each possible mention. In this paper, we propose a hybrid rule-neural coreference resolution system based on actor-critic learning, such that it can achieve better coreference performance by leveraging the advantages from both the heuristic rules and a neural conference model. This end-to-end system can also perform both mention detection and resolution by leveraging a joint training algorithm. We experiment on the BERT model to generate input span representations. Our model with the BERT span representation achieves the state-of-the-art performance among the models on the CoNLL-2012 Shared Task English Test Set.

comment: 11 pages, 3 figures. arXiv admin note: substantial text overlap with arXiv:2212.09028

☆ Rumour detection using graph neural network and oversampling in benchmark Twitter dataset

Shaswat Patel, Prince Bansal, Preeti Kaur

Recently, online social media has become a primary source for new information and misinformation or rumours. In the absence of an automatic rumour detection system the propagation of rumours has increased manifold leading to serious societal damages. In this work, we propose a novel method for building automatic rumour detection system by focusing on oversampling to alleviating the fundamental challenges of class imbalance in rumour detection task. Our oversampling method relies on contextualised data augmentation to generate synthetic samples for underrepresented classes in the dataset. The key idea exploits selection of tweets in a thread for augmentation which can be achieved by introducing a non-random selection criteria to focus the augmentation process on relevant tweets. Furthermore, we propose two graph neural networks(GNN) to model non-linear conversations on a thread. To enhance the tweet representations in our method we employed a custom feature selection technique based on state-of-the-art BERTweet model. Experiments of three publicly available datasets confirm that 1) our GNN models outperform the the current state-of-the-art classifiers by more than 20%(F1-score); 2) our oversampling technique increases the model performance by more than 9%;(F1-score) 3) focusing on relevant tweets for data augmentation via non-random selection criteria can further improve the results; and 4) our method has superior capabilities to detect rumours at very early stage.

☆ A Survey on Pretrained Language Models for Neural Code Intelligence

Yichen Xu, Yanqiao Zhu

As the complexity of modern software continues to escalate, software engineering has become an increasingly daunting and error-prone endeavor. In recent years, the field of Neural Code Intelligence (NCI) has emerged as a promising solution, leveraging the power of deep learning techniques to tackle analytical tasks on source code with the goal of improving programming efficiency and minimizing human errors within the software industry. Pretrained language models have become a dominant force in NCI research, consistently delivering state-of-the-art results across a wide range of tasks, including code summarization, generation, and translation. In this paper, we present a comprehensive survey of the NCI domain, including a thorough review of pretraining techniques, tasks, datasets, and model architectures. We hope this paper will serve as a bridge between the natural language and programming language communities, offering insights for future research in this rapidly evolving field.

comment: work in progress. 13 pages

☆ DOC: Improving Long Story Coherence With Detailed Outline Control

Kevin Yang, Dan Klein, Nanyun Peng, Yuandong Tian

We propose the Detailed Outline Control (DOC) framework for improving long-range plot coherence when automatically generating several-thousand-word-long stories. DOC consists of two complementary components: a detailed outliner and a detailed controller. The detailed outliner creates a more detailed, hierarchically structured outline, shifting creative burden from the main drafting procedure to the planning stage. The detailed controller ensures the more detailed outline is still respected during generation by controlling story passages to align with outline details. In human evaluations of automatically generated stories, DOC substantially outperforms a strong Re3 baseline (Yang et al., 2022) on plot coherence (22.5% absolute gain), outline relevance (28.2%), and interestingness (20.7%). Humans also judged DOC to be much more controllable in an interactive generation setting.

☆ Large Language Models Are Reasoning Teachers

Namgyu Ho, Laura Schmid, Se-Young Yun

Language models (LMs) have demonstrated remarkable performance on downstream tasks, using in-context exemplars or human instructions. Recent works have shown that chain-of-thought (CoT) prompting can elicit models to solve complex reasoning tasks, step-by-step. However, the efficacy of prompt-based CoT methods is restricted to very large LMs such as GPT-3 (175B), thus limiting deployability. In this paper, we revisit the fine-tuning approach to enable complex reasoning in smaller LMs, optimized to efficiently perform a specific task. We propose Fine-tune-CoT, a method that leverages the capabilities of very large LMs to generate reasoning samples and teach smaller models via fine-tuning. We evaluate our method on publicly available LMs across a wide range of complex tasks and model sizes. We find that Fine-tune-CoT enables substantial reasoning capability in small models, whereas previous prompt-based baselines exhibit near-random performance. Student models can even outperform the teacher in some tasks while reducing model size requirements by several orders of magnitude. We conduct extensive ablations and sample studies to understand the reasoning capabilities of student models. We also identify several important nuances that have been overlooked in concurrent fine-tuning works on CoT and address them in our analysis.

☆ An AI Dungeon Master's Guide: Learning to Converse and Guide with Intents and Theory-of-Mind in Dungeons and Dragons

Pei Zhou, Andrew Zhu, Jennifer Hu, Jay Pujara, Xiang Ren, Chris Callison-Burch, Yejin Choi, Prithviraj Ammanabrolu

We propose a novel task, G4C (Goal-driven Guidance Generation in Grounded Communication), for studying goal-driven and grounded natural language interactions. Specifically, we choose Dungeons and Dragons (D&D) -- a role-playing game consisting of multiple player characters and a Dungeon Master (DM) who collaborate to achieve a set of goals that are beneficial to the players -- as a testbed for this task. Here, each of the player characters is a student, with their own personas and abilities, and the DM is the teacher, an arbitrator of the rules of the world and responsible for assisting and guiding the students towards a global goal. We propose a theory-of-mind-inspired methodology for training such a DM with reinforcement learning (RL), where a DM: (1) learns to predict how the players will react to its utterances using a dataset of D&D dialogue transcripts; and (2) uses this prediction as a reward function providing feedback on how effective these utterances are at guiding the players towards a goal. Human and automated evaluations show that a DM trained with RL to generate guidance by incorporating a theory-of-mind of the players significantly improves the players' ability to achieve goals grounded in their shared world.

comment: 17 pages, 9 figures. Preprint, work in progress

☆ WeCheck: Strong Factual Consistency Checker via Weakly Supervised Learning

Wenhao Wu, Wei Li, Xinyan Xiao, Jiachen Liu, Sujian Li, Yajuan Lv

A crucial issue of current text generation models is that they often uncontrollably generate factually inconsistent text with respective of their inputs. Limited by the lack of annotated data, existing works in evaluating factual consistency directly transfer the reasoning ability of models trained on other data-rich upstream tasks like question answering (QA) and natural language inference (NLI) without any further adaptation. As a result, they perform poorly on the real generated text and are biased heavily by their single-source upstream tasks. To alleviate this problem, we propose a weakly supervised framework that aggregates multiple resources to train a precise and efficient factual metric, namely WeCheck. WeCheck first utilizes a generative model to accurately label a real generated sample by aggregating its weak labels, which are inferred from multiple resources. Then, we train the target metric model with the weak supervision while taking noises into consideration. Comprehensive experiments on a variety of tasks demonstrate the strong performance of WeCheck, which achieves a 3.4\% absolute improvement over previous state-of-the-art methods on TRUE benchmark on average.

☆ A Framework of Customer Review Analysis Using the Aspect-Based Opinion Mining Approach

Subhasis Dasgupta, Jaydip Sen

Opinion mining is the branch of computation that deals with opinions, appraisals, attitudes, and emotions of people and their different aspects. This field has attracted substantial research interest in recent years. Aspect-level (called aspect-based opinion mining) is often desired in practical applications as it provides detailed opinions or sentiments about different aspects of entities and entities themselves, which are usually required for action. Aspect extraction and entity extraction are thus two core tasks of aspect-based opinion mining. his paper has presented a framework of aspect-based opinion mining based on the concept of transfer learning. on real-world customer reviews available on the Amazon website. The model has yielded quite satisfactory results in its task of aspect-based opinion mining.

comment: This is the accepted version of the paper that has been presented and published in the 20th IEEE Conference, OCIT'22. The final published version is copyright-protected by the IEEE. The paper consists of 5 pages, and it includes 5 figures and 1 table

☆ An Augmentation Strategy for Visually Rich Documents

Jing Xie, James B. Wendt, Yichao Zhou, Seth Ebner, Sandeep Tata

Many business workflows require extracting important fields from form-like documents (e.g. bank statements, bills of lading, purchase orders, etc.). Recent techniques for automating this task work well only when trained with large datasets. In this work we propose a novel data augmentation technique to improve performance when training data is scarce, e.g. 10-250 documents. Our technique, which we call FieldSwap, works by swapping out the key phrases of a source field with the key phrases of a target field to generate new synthetic examples of the target field for use in training. We demonstrate that this approach can yield 1-7 F1 point improvements in extraction performance.

comment: 9 pages, 6 figures, 3 tables

☆ A Twitter BERT Approach for Offensive Language Detection in Marathi

Tanmay Chavan, Shantanu Patankar, Aditya Kane, Omkar Gokhale, Raviraj Joshi

Automated offensive language detection is essential in combating the spread of hate speech, particularly in social media. This paper describes our work on Offensive Language Identification in low resource Indic language Marathi. The problem is formulated as a text classification task to identify a tweet as offensive or non-offensive. We evaluate different mono-lingual and multi-lingual BERT models on this classification task, focusing on BERT models pre-trained with social media datasets. We compare the performance of MuRIL, MahaTweetBERT, MahaTweetBERT-Hateful, and MahaBERT on the HASOC 2022 test set. We also explore external data augmentation from other existing Marathi hate speech corpus HASOC 2021 and L3Cube-MahaHate. The MahaTweetBERT, a BERT model, pre-trained on Marathi tweets when fine-tuned on the combined dataset (HASOC 2021 + HASOC 2022 + MahaHate), outperforms all models with an F1 score of 98.43 on the HASOC 2022 test set. With this, we also provide a new state-of-the-art result on HASOC 2022 / MOLD v2 test set.

☆ Do language models have coherent mental models of everyday things?

Yuling Gu, Bhavana Dalvi Mishra, Peter Clark

When people think of everyday things like an "egg," they typically have a mental image associated with it. This commonsense knowledge helps us understand how these everyday things work and how to interact with them. For example, when someone tries to make a fried egg, they know that it has a shell and that it can be cracked open to reveal the egg white and yolk inside. However, if a system does not have a coherent picture of such everyday things, thinking that the egg yolk surrounds the shell, then it might have to resort to ridiculous approaches such as trying to scrape the egg yolk off the shell into the pan. Do language models have a coherent picture of such everyday things? To investigate this, we propose a benchmark dataset consisting of 100 everyday things, their parts, and the relationships between these parts. We observe that state-of-the-art pre-trained language models (LMs) like GPT-3 and Macaw have fragments of knowledge about these entities, but they fail to produce consistent parts mental models. We propose a simple extension to these LMs where we apply a constraint satisfaction layer on top of raw predictions from LMs to produce more consistent and accurate parts mental models of everyday things.

☆ When Federated Learning Meets Pre-trained Language Models' Parameter-Efficient Tuning Methods

Zhuo Zhang, Yuanhang Yang, Yong Dai, Lizhen Qu, Zenglin Xu

With increasing privacy concerns on data, recent studies have made significant progress using federated learning (FL) on privacy-sensitive natural language processing (NLP) tasks. Much literature suggests fully fine-tuning pre-trained language models (PLMs) in the FL paradigm can mitigate the data heterogeneity problem and close the performance gap with centralized training. However, large PLMs bring the curse of prohibitive communication overhead and local model adaptation costs for the FL system. To this end, we introduce various parameter-efficient tuning (PETuning) methods into federated learning. Specifically, we provide a holistic empirical study of representative PLMs tuning methods in FL. The experimental results cover the analysis of data heterogeneity levels, data scales, and different FL scenarios. Overall communication overhead can be significantly reduced by locally tuning and globally aggregating lightweight model parameters while maintaining acceptable performance in various FL settings. To facilitate the research of PETuning in FL, we also develop a federated tuning framework FedPETuning, which allows practitioners to exploit different PETuning methods under the FL training paradigm conveniently. The source code is available at \url{https://github.com/iezhuozhuo/FedETuning/tree/deltaTuning}.

☆ On the Blind Spots of Model-Based Evaluation Metrics for Text Generation

Tianxing He, Jingyu Zhang, Tianle Wang, Sachin Kumar, Kyunghyun Cho, James Glass, Yulia Tsvetkov

In this work, we explore a useful but often neglected methodology for robustness analysis of text generation evaluation metrics: stress tests with synthetic data. Basically, we design and synthesize a wide range of potential errors and check whether they result in a commensurate drop in the metric scores. We examine a range of recently proposed evaluation metrics based on pretrained language models, for the tasks of open-ended generation, translation, and summarization. Our experiments reveal interesting insensitivities, biases, or even loopholes in existing metrics. For example, we find that BERTScore ignores truncation errors in summarization, and MAUVE (built on top of GPT-2) is insensitive to errors at the beginning of generations. Further, we investigate the reasons behind these blind spots and suggest practical workarounds for a more reliable evaluation of text generation.

☆ When Do Decompositions Help for Machine Reading?

Kangda Wei, Dawn Lawrie, Benjamin Van Durme, Yunmo Chen, Orion Weller

Answering complex questions often requires multi-step reasoning in order to obtain the final answer. Most research into decompositions of complex questions involves open-domain systems, which have shown success in using these decompositions for improved retrieval. In the machine reading setting, however, work to understand when decompositions are helpful is understudied. We conduct experiments on decompositions in machine reading to unify recent work in this space, using a range of models and datasets. We find that decompositions can be helpful in the few-shot case, giving several points of improvement in exact match scores. However, we also show that when models are given access to datasets with around a few hundred or more examples, decompositions are not helpful (and can actually be detrimental). Thus, our analysis implies that models can learn decompositions implicitly even with limited data.

☆ DIONYSUS: A Pre-trained Model for Low-Resource Dialogue Summarization

Yu Li, Baolin Peng, Pengcheng He, Michel Galley, Zhou Yu, Jianfeng Gao

Dialogue summarization has recently garnered significant attention due to its wide range of applications. However, existing methods for summarizing dialogues are suboptimal because they do not take into account the inherent structure of dialogue and rely heavily on labeled data, which can lead to poor performance in new domains. In this work, we propose DIONYSUS (dynamic input optimization in pre-training for dialogue summarization), a pre-trained encoder-decoder model for summarizing dialogues in any new domain. To pre-train DIONYSUS, we create two pseudo summaries for each dialogue example: one is produced by a fine-tuned summarization model, and the other is a collection of dialogue turns that convey important information. We then choose one of these pseudo summaries based on the difference in information distribution across different types of dialogues. This selected pseudo summary serves as the objective for pre-training DIONYSUS using a self-supervised approach on a large dialogue corpus. Our experiments show that DIONYSUS outperforms existing methods on six datasets, as demonstrated by its ROUGE scores in zero-shot and few-shot settings.

☆ Benchmarking Spatial Relationships in Text-to-Image Generation

Tejas Gokhale, Hamid Palangi, Besmira Nushi, Vibhav Vineet, Eric Horvitz, Ece Kamar, Chitta Baral, Yezhou Yang

Spatial understanding is a fundamental aspect of computer vision and integral for human-level reasoning about images, making it an important component for grounded language understanding. While recent large-scale text-to-image synthesis (T2I) models have shown unprecedented improvements in photorealism, it is unclear whether they have reliable spatial understanding capabilities. We investigate the ability of T2I models to generate correct spatial relationships among objects and present VISOR, an evaluation metric that captures how accurately the spatial relationship described in text is generated in the image. To benchmark existing models, we introduce a large-scale challenge dataset SR2D that contains sentences describing two objects and the spatial relationship between them. We construct and harness an automated evaluation pipeline that employs computer vision to recognize objects and their spatial relationships, and we employ it in a large-scale evaluation of T2I models. Our experiments reveal a surprising finding that, although recent state-of-the-art T2I models exhibit high image quality, they are severely limited in their ability to generate multiple objects or the specified spatial relations such as left/right/above/below. Our analyses demonstrate several biases and artifacts of T2I models such as the difficulty with generating multiple objects, a bias towards generating the first object mentioned, spatially inconsistent outputs for equivalent relationships, and a correlation between object co-occurrence and spatial understanding capabilities. We conduct a human study that shows the alignment between VISOR and human judgment about spatial understanding. We offer the SR2D dataset and the VISOR metric to the community in support of T2I spatial reasoning research.

comment: preprint

☆ DocAsRef: A Pilot Empirical Study on Repurposing Reference-Based Summary Quality Metrics Reference-Freely

Forrest Sheng Bao, Ruixuan Tu, Ge Luo

Summary quality assessment metrics have two categories: reference-based and reference-free. Reference-based metrics are theoretically more accurate but are limited by the availability and quality of the human-written references, which are both difficulty to ensure. This inspires the development of reference-free metrics, which are independent from human-written references, in the past few years. However, existing reference-free metrics cannot be both zero-shot and accurate. In this paper, we propose a zero-shot but accurate reference-free approach in a sneaky way: feeding documents, based upon which summaries generated, as references into reference-based metrics. Experimental results show that this zero-shot approach can give us the best-performing reference-free metrics on nearly all aspects on several recently-released datasets, even beating reference-free metrics specifically trained for this task sometimes. We further investigate what reference-based metrics can benefit from such repurposing and whether our additional tweaks help.

comment: a pilot study

☆ Language Modeling with Latent Situations

Belinda Z. Li, Maxwell Nye, Jacob Andreas

Language models (LMs) often generate incoherent outputs: they refer to events and entity states that are incompatible with the state of the world described in their inputs. We introduce SituationSupervision, a family of approaches for improving coherence in LMs by training them to construct and condition on explicit representations of entities and their states. SituationSupervision has two components: an auxiliary situation modeling task that trains models to predict state representations in context, and a latent state inference procedure that imputes these states from partially annotated training data. SituationSupervision can be applied to both fine-tuning (by supervising LMs to encode state variables in their hidden representations) and prompting (by inducing LMs to interleave textual descriptions of entity states with output text). In both cases, SituationSupervision requires only a small number of state annotations to produce major coherence improvements (between 4-11%), showing that standard LMs can be sample-efficiently trained to model not just language but the situations it describes.

comment: 13 pages, 3 figures, 7 tables

☆ PLUE: Language Understanding Evaluation Benchmark for Privacy Policies in English

Jianfeng Chi, Wasi Uddin Ahmad, Yuan Tian, Kai-Wei Chang

Privacy policies provide individuals with information about their rights and how their personal information is handled. Natural language understanding (NLU) technologies can support individuals and practitioners to understand better privacy practices described in lengthy and complex documents. However, existing efforts that use NLU technologies are limited by processing the language in a way exclusive to a single task focusing on certain privacy practices. To this end, we introduce the Privacy Policy Language Understanding Evaluation (PLUE) benchmark, a multi-task benchmark for evaluating the privacy policy language understanding across various tasks. We also collect a large corpus of privacy policies to enable privacy policy domain-specific language model pre-training. We demonstrate that domain-specific pre-training offers performance improvements across all tasks. We release the benchmark to encourage future research in this domain.

☆ Enhancing Task Bot Engagement with Synthesized Open-Domain Dialog

Miaoran Li, Baolin Peng, Michel Galley, Jianfeng Gao, Zhu Zhang

Many efforts have been made to construct dialog systems for different types of conversations, such as task-oriented dialog (TOD) and open-domain dialog (ODD). To better mimic human-level conversations that usually fuse various dialog modes, it is essential to build a system that can effectively handle both TOD and ODD and access different knowledge sources. To address the lack of available data for the fused task, we propose a framework for automatically generating dialogues that combine knowledge-grounded ODDs and TODs in various settings. Additionally, we introduce a unified model PivotBot that is capable of appropriately adopting TOD and ODD modes and accessing different knowledge sources in order to effectively tackle the fused task. Evaluation results demonstrate the superior ability of the proposed model to switch seamlessly between TOD and ODD tasks.

☆ CoCoMIC: Code Completion By Jointly Modeling In-file and Cross-file Context

Yangruibo Ding, Zijian Wang, Wasi Uddin Ahmad, Murali Krishna Ramanathan, Ramesh Nallapati, Parminder Bhatia, Dan Roth, Bing Xiang

While pre-trained language models (LM) for code have achieved great success in code completion, they generate code conditioned only on the contents within the file, i.e., in-file context, but ignore the rich semantics in other files within the same project, i.e., cross-file context, a critical source of information that is especially useful in modern modular software development. Such overlooking constrains code language models' capacity in code completion, leading to unexpected behaviors such as generating hallucinated class member functions or function calls with unexpected arguments. In this work, we develop a cross-file context finder tool, CCFINDER, that effectively locates and retrieves the most relevant cross-file context. We propose CoCoMIC, a framework that incorporates cross-file context to learn the in-file and cross-file context jointly on top of pretrained code LMs. CoCoMIC successfully improves the existing code LM with a 19.30% relative increase in exact match and a 15.41% relative increase in identifier matching for code completion when the cross-file context is provided.

☆ (QA)$^2$: Question Answering with Questionable Assumptions

Najoung Kim, Phu Mon Htut, Samuel R. Bowman, Jackson Petty

Naturally-occurring information-seeking questions often contain questionable assumptions -- assumptions that are false or unverifiable. Questions containing questionable assumptions are challenging because they require a distinct answer strategy that deviates from typical answers to information-seeking questions. For instance, the question "When did Marie Curie discover Uranium?" cannot be answered as a typical when question without addressing the false assumption "Marie Curie discovered Uranium". In this work, we propose (QA)$^2$ (Question Answering with Questionable Assumptions), an open-domain evaluation dataset consisting of naturally-occurring search engine queries that may or may not contain questionable assumptions. To be successful on (QA)$^2$, systems must be able to detect questionable assumptions and also be able to produce adequate responses for both typical information-seeking questions and ones with questionable assumptions. We find that current models do struggle with handling questionable assumptions -- the best performing model achieves 59% human rater acceptability on abstractive QA with (QA)$^2$ questions, leaving substantial headroom for progress.

comment: Preprint

☆ Defending Against Poisoning Attacks in Open-Domain Question Answering

Orion Weller, Aleem Khan, Nathaniel Weir, Dawn Lawrie, Benjamin Van Durme

Recent work in open-domain question answering (ODQA) has shown that adversarial poisoning of the input contexts can cause large drops in accuracy for production systems. However, little to no work has proposed methods to defend against these attacks. To do so, we introduce a new method that uses query augmentation to search for a diverse set of retrieved passages that could answer the original question. We integrate these new passages into the model through the design of a novel confidence method, comparing the predicted answer to its appearance in the retrieved contexts (what we call Confidence from Answer Redundancy, e.g. CAR). Together these methods allow for a simple but effective way to defend against poisoning attacks and provide gains of 5-20% exact match across varying levels of data poisoning.

☆ Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What Matters

Boshi Wang, Sewon Min, Xiang Deng, Jiaming Shen, You Wu, Luke Zettlemoyer, Huan Sun

Chain-of-Thought (CoT) prompting can dramatically improve the multi-step reasoning abilities of large language models (LLMs). CoT explicitly encourages the LLM to generate intermediate rationales for solving a problem, by providing a series of reasoning steps in the demonstrations. Despite its success, there is still little understanding of what makes CoT prompting effective and which aspects of the demonstrated reasoning steps contribute to its performance. In this paper, we show that CoT reasoning is possible even with invalid demonstrations - prompting with invalid reasoning steps can achieve over 80-90% of the performance obtained using CoT under various metrics, while still generating coherent lines of reasoning during inference. Further experiments show that other aspects of the rationales, such as being relevant to the query and correctly ordering the reasoning steps, are much more important for effective CoT reasoning. Overall, these findings both deepen our understanding of CoT prompting, and open up new questions regarding LLMs' capability to learn to reason in context.

comment: Our code and model input/output will be available at https://github.com/sunlab-osu/Understanding-CoT

☆ Towards Robustness of Text-to-SQL Models Against Natural and Realistic Adversarial Table Perturbation ACL 2022

Xinyu Pi, Bing Wang, Yan Gao, Jiaqi Guo, Zhoujun Li, Jian-Guang Lou

The robustness of Text-to-SQL parsers against adversarial perturbations plays a crucial role in delivering highly reliable applications. Previous studies along this line primarily focused on perturbations in the natural language question side, neglecting the variability of tables. Motivated by this, we propose the Adversarial Table Perturbation (ATP) as a new attacking paradigm to measure the robustness of Text-to-SQL models. Following this proposition, we curate ADVETA, the first robustness evaluation benchmark featuring natural and realistic ATPs. All tested state-of-the-art models experience dramatic performance drops on ADVETA, revealing models' vulnerability in real-world practices. To defend against ATP, we build a systematic adversarial training example generation framework tailored for better contextualization of tabular data. Experiments show that our approach not only brings the best robustness improvement against table-side perturbations but also substantially empowers models against NL-side perturbations. We release our benchmark and code at: https://github.com/microsoft/ContextualSP.

comment: Accepted by ACL 2022 (Oral)

☆ Joint Speech Transcription and Translation: Pseudo-Labeling with Out-of-Distribution Data

Mozhdeh Gheini, Tatiana Likhomanenko, Matthias Sperber, Hendra Setiawan

Self-training has been shown to be helpful in addressing data scarcity for many domains, including vision, speech, and language. Specifically, self-training, or pseudo-labeling, labels unsupervised data and adds that to the training pool. In this work, we investigate and use pseudo-labeling for a recently proposed novel setup: joint transcription and translation of speech, which suffers from an absence of sufficient data resources. We show that under such data-deficient circumstances, the unlabeled data can significantly vary in domain from the supervised data, which results in pseudo-label quality degradation. We investigate two categories of remedies that require no additional supervision and target the domain mismatch: pseudo-label filtering and data augmentation. We show that pseudo-label analysis and processing as such results in additional gains on top of the vanilla pseudo-labeling setup resulting in total improvements of up to 0.6% absolute WER and 2.2 BLEU points.

☆ On Improving Summarization Factual Consistency from Natural Language Feedback

Yixin Liu, Budhaditya Deb, Milagro Teruel, Aaron Halfaker, Dragomir Radev, Ahmed H. Awadallah

Despite the recent progress in language generation models, their outputs may not always meet user expectations. In this work, we study whether informational feedback in natural language can be leveraged to improve generation quality and user preference alignment. To this end, we consider factual consistency in summarization, the quality that the summary should only contain information supported by the input documents, for user preference alignment. We collect a high-quality dataset, DeFacto, containing human demonstrations and informational feedback in natural language consisting of corrective instructions, edited summaries, and explanations with respect to the factual consistency of the summary. Using our dataset, we study two natural language generation tasks: 1) editing a summary using the human feedback, and 2) generating human feedback from the original summary. Using the two tasks, we further evaluate if models can automatically correct factual inconsistencies in generated summaries. We show that the human-edited summaries we collected are more factually consistent, and pre-trained language models can leverage our dataset to improve the factual consistency of original system-generated summaries in our proposed generation tasks. We make the DeFacto dataset publicly available at https://github.com/microsoft/DeFacto.

☆ BUMP: A Benchmark of Unfaithful Minimal Pairs for Meta-Evaluation of Faithfulness Metrics

Liang Ma, Shuyang Cao, Robert L. Logan IV, Di Lu, Shihao Ran, Ke Zhang, Joel Tetreault, Aoife Cahill, Alejandro Jaimes

The proliferation of automatic faithfulness metrics for summarization has produced a need for benchmarks to evaluate them. While existing benchmarks measure the correlation with human judgements of faithfulness on model-generated summaries, they are insufficient for diagnosing whether metrics are: 1) consistent, i.e., decrease as errors are introduced into a summary, 2) effective on human-written texts, and 3) sensitive to different error types (as summaries can contain multiple errors). To address these needs, we present a benchmark of unfaithful minimal pairs (BUMP), a dataset of 889 human-written, minimally different summary pairs, where a single error (from an ontology of 7 types) is introduced to a summary from the CNN/DailyMail dataset to produce an unfaithful summary. We find BUMP complements existing benchmarks in a number of ways: 1) the summaries in BUMP are harder to discriminate and less probable under SOTA summarization models, 2) BUMP enables measuring the consistency of metrics, and reveals that the most discriminative metrics tend not to be the most consistent, 3) BUMP enables the measurement of metrics' performance on individual error types and highlights areas of weakness for future work.

☆ Future Sight: Dynamic Story Generation with Large Pretrained Language Models

Brian D. Zimmerman, Gaurav Sahu, Olga Vechtomova

Recent advances in deep learning research, such as transformers, have bolstered the ability for automated agents to generate creative texts similar to those that a human would write. By default, transformer decoders can only generate new text with respect to previously generated text. The output distribution of candidate tokens at any position is conditioned on previously selected tokens using a self-attention mechanism to emulate the property of autoregression. This is inherently limiting for tasks such as controllable story generation where it may be necessary to condition on future plot events when writing a story. In this work, we propose Future Sight, a method for finetuning a pretrained generative transformer on the task of future conditioning. Transformer decoders are typically pretrained on the task of completing a context, one token at a time, by means of self-attention. Future Sight additionally enables a decoder to attend to an encoded future plot event. This motivates the decoder to expand on the context in a way that logically concludes with the provided future. During inference, the future plot event can be written by a human author to steer the narrative being generated in a certain direction. We evaluate the efficacy of our approach on a story generation task with human evaluators.

comment: 9 pages, 1 figure, 4 tables

☆ Dialog2API: Task-Oriented Dialogue with API Description and Example Programs

Raphael Shu, Elman Mansimov, Tamer Alkhouli, Nikolaos Pappas, Salvatore Romeo, Arshit Gupta, Saab Mansour, Yi Zhang, Dan Roth

Functionality and dialogue experience are two important factors of task-oriented dialogue systems. Conventional approaches with closed schema (e.g., conversational semantic parsing) often fail as both the functionality and dialogue experience are strongly constrained by the underlying schema. We introduce a new paradigm for task-oriented dialogue - Dialog2API - to greatly expand the functionality and provide seamless dialogue experience. The conversational model interacts with the environment by generating and executing programs triggering a set of pre-defined APIs. The model also manages the dialogue policy and interact with the user through generating appropriate natural language responses. By allowing generating free-form programs, Dialog2API supports composite goals by combining different APIs, whereas unrestricted program revision provides natural and robust dialogue experience. To facilitate Dialog2API, the core model is provided with API documents, an execution environment and optionally some example dialogues annotated with programs. We propose an approach tailored for the Dialog2API, where the dialogue states are represented by a stack of programs, with most recently mentioned program on the top of the stack. Dialog2API can work with many application scenarios such as software automation and customer service. In this paper, we construct a dataset for AWS S3 APIs and present evaluation results of in-context learning baselines.

☆ AnyTOD: A Programmable Task-Oriented Dialog System

Jeffrey Zhao, Yuan Cao, Raghav Gupta, Harrison Lee, Abhinav Rastogi, Mingqiu Wang, Hagen Soltau, Izhak Shafran, Yonghui Wu

We propose AnyTOD, an end-to-end task-oriented dialog (TOD) system with zero-shot capability for unseen tasks. We view TOD as a program executed by a language model (LM), where program logic and ontology is provided by a designer in the form of a schema. To enable generalization onto unseen schemas and programs without prior training, AnyTOD adopts a neuro-symbolic approach. A neural LM keeps track of events that occur during a conversation, and a symbolic program implementing the dialog policy is executed to recommend next actions AnyTOD should take. This approach drastically reduces data annotation and model training requirements, addressing a long-standing challenge in TOD research: rapidly adapting a TOD system to unseen tasks and domains. We demonstrate state-of-the-art results on the STAR and ABCD benchmarks, as well as AnyTOD's strong zero-shot transfer capability in low-resource settings. In addition, we release STARv2, an updated version of the STAR dataset with richer data annotations, for benchmarking zero-shot end-to-end TOD models.

☆ Improving the Robustness of Summarization Models by Detecting and Removing Input Noise

Kundan Krishna, Yao Zhao, Jie Ren, Balaji Lakshminarayanan, Jiaming Luo, Mohammad Saleh, Peter J. Liu

The evaluation of abstractive summarization models typically uses test data that is identically distributed as training data. In real-world practice, documents to be summarized may contain input noise caused by text extraction artifacts or data pipeline bugs. The robustness of model performance under distribution shift caused by such noise is relatively under-studied. We present a large empirical study quantifying the sometimes severe loss in performance (up to 12 ROUGE-1 points) from different types of input noise for a range of datasets and model sizes. We then propose a light-weight method for detecting and removing such noise in the input during model inference without requiring any extra training, auxiliary models, or even prior knowledge of the type of noise. Our proposed approach effectively mitigates the loss in performance, recovering a large fraction of the performance drop, sometimes as large as 11 ROUGE-1 points.

☆ Generation-Augmented Query Expansion For Code Retrieval

Dong Li, Yelong Shen, Ruoming Jin, Yi Mao, Kuan Wang, Weizhu Chen

Pre-trained language models have achieved promising success in code retrieval tasks, where a natural language documentation query is given to find the most relevant existing code snippet. However, existing models focus only on optimizing the documentation code pairs by embedding them into latent space, without the association of external knowledge. In this paper, we propose a generation-augmented query expansion framework. Inspired by the human retrieval process - sketching an answer before searching, in this work, we utilize the powerful code generation model to benefit the code retrieval task. Specifically, we demonstrate that rather than merely retrieving the target code snippet according to the documentation query, it would be helpful to augment the documentation query with its generation counterpart - generated code snippets from the code generation model. To the best of our knowledge, this is the first attempt that leverages the code generation model to enhance the code retrieval task. We achieve new state-of-the-art results on the CodeSearchNet benchmark and surpass the baselines significantly.

☆ METEOR Guided Divergence for Video Captioning

Daniel Lukas Rothenpieler, Shahin Amiriparian

Automatic video captioning aims for a holistic visual scene understanding. It requires a mechanism for capturing temporal context in video frames and the ability to comprehend the actions and associations of objects in a given timeframe. Such a system should additionally learn to abstract video sequences into sensible representations as well as to generate natural written language. While the majority of captioning models focus solely on the visual inputs, little attention has been paid to the audiovisual modality. To tackle this issue, we propose a novel two-fold approach. First, we implement a reward-guided KL Divergence to train a video captioning model which is resilient towards token permutations. Second, we utilise a Bi-Modal Hierarchical Reinforcement Learning (BMHRL) Transformer architecture to capture long-term temporal dependencies of the input data as a foundation for our hierarchical captioning module. Using our BMHRL, we show the suitability of the HRL agent in the generation of content-complete and grammatically sound sentences by achieving $4.91$, $2.23$, and $10.80$ in BLEU3, BLEU4, and METEOR scores, respectively on the ActivityNet Captions dataset. Finally, we make our BMHRL framework and trained models publicly available for users and developers at https://github.com/d-rothen/bmhrl.

☆ Understanding Stereotypes in Language Models: Towards Robust Measurement and Zero-Shot Debiasing

Justus Mattern, Zhijing Jin, Mrinmaya Sachan, Rada Mihalcea, Bernhard Schölkopf

Generated texts from large pretrained language models have been shown to exhibit a variety of harmful, human-like biases about various demographics. These findings prompted large efforts aiming to understand and measure such effects, with the goal of providing benchmarks that can guide the development of techniques mitigating these stereotypical associations. However, as recent research has pointed out, the current benchmarks lack a robust experimental setup, consequently hindering the inference of meaningful conclusions from their evaluation metrics. In this paper, we extend these arguments and demonstrate that existing techniques and benchmarks aiming to measure stereotypes tend to be inaccurate and consist of a high degree of experimental noise that severely limits the knowledge we can gain from benchmarking language models based on them. Accordingly, we propose a new framework for robustly measuring and quantifying biases exhibited by generative language models. Finally, we use this framework to investigate GPT-3's occupational gender bias and propose prompting techniques for mitigating these biases without the need for fine-tuning.

☆ In-context Learning Distillation: Transferring Few-shot Learning Ability of Pre-trained Language Models

Yukun Huang, Yanda Chen, Zhou Yu, Kathleen McKeown

Given the success with in-context learning of large pre-trained language models, we introduce in-context learning distillation to transfer in-context few-shot learning ability from large models to smaller models. We propose to combine in-context learning objectives with language modeling objectives to distill both the ability to read in-context examples and task knowledge to the smaller models. We perform in-context learning distillation under two different few-shot learning paradigms: Meta In-context Tuning (Meta-ICT) and Multitask In-context Tuning (Multitask-ICT). Multitask-ICT performs better on multitask few-shot learning but also requires more computation than Meta-ICT. Our method shows consistent improvements for both Meta-ICT and Multitask-ICT on two benchmarks: LAMA and CrossFit. Our extensive experiments and analysis reveal that in-context learning objectives and language modeling objectives are complementary under the Multitask-ICT paradigm. In-context learning objectives achieve the best performance when combined with language modeling objectives.

☆ KronA: Parameter Efficient Tuning with Kronecker Adapter

Ali Edalati, Marzieh Tahaei, Ivan Kobyzev, Vahid Partovi Nia, James J. Clark, Mehdi Rezagholizadeh

Fine-tuning a Pre-trained Language Model (PLM) on a specific downstream task has been a well-known paradigm in Natural Language Processing. However, with the ever-growing size of PLMs, training the entire model on several downstream tasks becomes very expensive and resource-hungry. Recently, different Parameter Efficient Tuning (PET) techniques are proposed to improve the efficiency of fine-tuning PLMs. One popular category of PET methods is the low-rank adaptation methods which insert learnable truncated SVD modules into the original model either sequentially or in parallel. However, low-rank decomposition suffers from limited representation power. In this work, we address this problem using the Kronecker product instead of the low-rank representation. We introduce KronA, a Kronecker product-based adapter module for efficient fine-tuning of Transformer-based PLMs. We apply the proposed methods for fine-tuning T5 on the GLUE benchmark to show that incorporating the Kronecker-based modules can outperform state-of-the-art PET methods.

☆ mFACE: Multilingual Summarization with Factual Consistency Evaluation

Roee Aharoni, Shashi Narayan, Joshua Maynez, Jonathan Herzig, Elizabeth Clark, Mirella Lapata

Abstractive summarization has enjoyed renewed interest in recent years, thanks to pre-trained language models and the availability of large-scale datasets. Despite promising results, current models still suffer from generating factually inconsistent summaries, reducing their utility for real-world application. Several recent efforts attempt to address this by devising models that automatically detect factual inconsistencies in machine generated summaries. However, they focus exclusively on English, a language with abundant resources. In this work, we leverage factual consistency evaluation models to improve multilingual summarization. We explore two intuitive approaches to mitigate hallucinations based on the signal provided by a multilingual NLI model, namely data filtering and controlled generation. Experimental results in the 45 languages from the XLSum dataset show gains over strong baselines in both automatic and human evaluation.

comment: 20 pages

☆ Ontologically Faithful Generation of Non-Player Character Dialogues

Nathaniel Weir, Ryan Thomas, Randolph D'Amore, Kellie Hill, Benjamin Van Durme, Harsh Jhamtani

We introduce a language generation task grounded in a popular video game environment. KNUDGE (KNowledge Constrained User-NPC Dialogue GEneration) involves generating dialogue trees conditioned on an ontology captured in natural language passages providing quest and entity specifications. KNUDGE is constructed from side quest dialogues drawn directly from game data of Obsidian Entertainment's The Outer Worlds, leading to real-world complexities in generation: (1) dialogues are branching trees as opposed to linear chains of utterances; (2) utterances must remain faithful to the game lore--character personas, backstories, and entity relationships; and (3) a dialogue must accurately reveal new quest-related details to the human player. We report results for supervised and in-context learning techniques, finding there is significant room for future work on creating realistic game-quality dialogues.

☆ Task Ambiguity in Humans and Language Models

Alex Tamkin, Kunal Handa, Avash Shrestha, Noah Goodman

Language models have recently achieved strong performance across a wide range of NLP benchmarks. However, unlike benchmarks, real world tasks are often poorly specified, and agents must deduce the user's intended behavior from a combination of context, instructions, and examples. We investigate how both humans and models behave in the face of such task ambiguity by proposing AmbiBench, a new benchmark of six ambiguously-specified classification tasks. We evaluate humans and models on AmbiBench by seeing how well they identify the intended task using 1) instructions with varying degrees of ambiguity, and 2) different numbers of labeled examples. We find that the combination of model scaling (to 175B parameters) and training with human feedback data enables models to approach or exceed the accuracy of human participants across tasks, but that either one alone is not sufficient. In addition, we show how to dramatically improve the accuracy of language models trained without large-scale human feedback training by finetuning on a small number of ambiguous in-context examples, providing a promising direction for teaching models to generalize well in the face of ambiguity.

☆ A survey on text generation using generative adversarial networks

Gustavo Henrique de Rosa, João Paulo Papa

This work presents a thorough review concerning recent studies and text generation advancements using Generative Adversarial Networks. The usage of adversarial learning for text generation is promising as it provides alternatives to generate the so-called "natural" language. Nevertheless, adversarial text generation is not a simple task as its foremost architecture, the Generative Adversarial Networks, were designed to cope with continuous information (image) instead of discrete data (text). Thus, most works are based on three possible options, i.e., Gumbel-Softmax differentiation, Reinforcement Learning, and modified training objectives. All alternatives are reviewed in this survey as they present the most recent approaches for generating text using adversarial-based techniques. The selected works were taken from renowned databases, such as Science Direct, IEEEXplore, Springer, Association for Computing Machinery, and arXiv, whereas each selected work has been critically analyzed and assessed to present its objective, methodology, and experimental results.

☆ Smooth Sailing: Improving Active Learning for Pre-trained Language Models with Representation Smoothness Analysis

Josip Jukić, Jan Šnajder

Developed as a solution to a practical need, active learning (AL) methods aim to reduce label complexity and the annotations costs in supervised learning. While recent work has demonstrated the benefit of using AL in combination with large pre-trained language models (PLMs), it has often overlooked the practical challenges that hinder the feasibility of AL in realistic settings. We address these challenges by leveraging representation smoothness analysis to improve the effectiveness of AL. We develop an early stopping technique that does not require a validation set -- often unavailable in realistic AL settings -- and observe significant improvements across multiple datasets and AL methods. Additionally, we find that task adaptation improves AL, whereas standard short fine-tuning in AL does not provide improvements over random sampling. Our work establishes the usefulness of representation smoothness analysis in AL and presents an AL stopping criterion that reduces label complexity.

☆ Trustworthy Social Bias Measurement

Rishi Bommasani, Percy Liang

How do we design measures of social bias that we trust? While prior work has introduced several measures, no measure has gained widespread trust: instead, mounting evidence argues we should distrust these measures. In this work, we design bias measures that warrant trust based on the cross-disciplinary theory of measurement modeling. To combat the frequently fuzzy treatment of social bias in NLP, we explicitly define social bias, grounded in principles drawn from social science research. We operationalize our definition by proposing a general bias measurement framework DivDist, which we use to instantiate 5 concrete bias measures. To validate our measures, we propose a rigorous testing protocol with 8 testing criteria (e.g. predictive validity: do measures predict biases in US employment?). Through our testing, we demonstrate considerable evidence to trust our measures, showing they overcome conceptual, technical, and empirical deficiencies present in prior measures.

☆ Evaluation for Change

Rishi Bommasani

Evaluation is the central means for assessing, understanding, and communicating about NLP models. In this position paper, we argue evaluation should be more than that: it is a force for driving change, carrying a sociological and political character beyond its technical dimensions. As a force, evaluation's power arises from its adoption: under our view, evaluation succeeds when it achieves the desired change in the field. Further, by framing evaluation as a force, we consider how it competes with other forces. Under our analysis, we conjecture that the current trajectory of NLP suggests evaluation's power is waning, in spite of its potential for realizing more pluralistic ambitions in the field. We conclude by discussing the legitimacy of this power, who acquires this power and how it distributes. Ultimately, we hope the research community will more aggressively harness evaluation for change.

comment: Exploratory position paper

♻ ☆ Evaluating Human-Language Model Interaction

Mina Lee, Megha Srivastava, Amelia Hardy, John Thickstun, Esin Durmus, Ashwin Paranjape, Ines Gerard-Ursin, Xiang Lisa Li, Faisal Ladhak, Frieda Rong, Rose E. Wang, Minae Kwon, Joon Sung Park, Hancheng Cao, Tony Lee, Rishi Bommasani, Michael Bernstein, Percy Liang

Many real-world applications of language models (LMs), such as code autocomplete and writing assistance, involve human-LM interaction. However, the main LM benchmarks are non-interactive in that a system produces output without human involvement. To evaluate human-LM interaction, we develop a new framework, Human-AI Language-based Interaction Evaluation (HALIE), that expands non-interactive evaluation along three dimensions, capturing (i) the interactive process, not only the final output; (ii) the first-person subjective experience, not just a third-party assessment; and (iii) notions of preference beyond quality. We then design five tasks ranging from goal-oriented to open-ended to capture different forms of interaction. On four state-of-the-art LMs (three variants of OpenAI's GPT-3 and AI21's J1-Jumbo), we find that non-interactive performance does not always result in better human-LM interaction and that first-person and third-party metrics can diverge, suggesting the importance of examining the nuances of human-LM interaction.

♻ ☆ BertNet: Harvesting Knowledge Graphs from Pretrained Language Models

Shibo Hao, Bowen Tan, Kaiwen Tang, Bin Ni, Hengzhe Zhang, Eric P Xing, Zhiting Hu

Symbolic knowledge graphs (KGs) have been constructed either by expensive human crowdsourcing or with complex text mining pipelines. The emerging large pretrained language models (LMs), such as Bert, have shown to implicitly encode massive knowledge which can be queried with properly designed prompts. However, compared to the explicit KGs, the implict knowledge in the black-box LMs is often difficult to access or edit and lacks explainability. In this work, we aim at harvesting symbolic KGs from the LMs, and propose a new framework for automatic KG construction empowered by the neural LMs' flexibility and scalability. Compared to prior works that often rely on large human annotated data or existing massive KGs, our approach requires only the minimal definition of relations as inputs, and hence is suitable for extracting knowledge of rich new relations that are instantly assigned and not available before. The framework automatically generates diverse prompts, and performs efficient knowledge search within a given LM for consistent outputs. The knowledge harvested with our approach shows competitive quality, diversity, and novelty. As a result, we derive from diverse LMs a family of new KGs (e.g., BertNet and RoBERTaNet) that contain a richer set of relations, including some complex ones (e.g., "A is capable of but not good at B") that cannot be extracted with previous methods. Besides, the resulting KGs also serve as a vehicle to interpret the respective source LMs, leading to new insights into the varying knowledge capability of different LMs.

comment: Code available at https://github.com/tanyuqian/knowledge-harvest-from-lms

♻ ☆ Mind the Knowledge Gap: A Survey of Knowledge-enhanced Dialogue Systems

Sagi Shaier, Lawrence Hunter, Katharina Kann

Many dialogue systems (DSs) lack characteristics humans have, such as emotion perception, factuality, and informativeness. Enhancing DSs with knowledge alleviates this problem, but, as many ways of doing so exist, keeping track of all proposed methods is difficult. Here, we present the first survey of knowledge-enhanced DSs. We define three categories of systems - internal, external, and hybrid - based on the knowledge they use. We survey the motivation for enhancing DSs with knowledge, used datasets, and methods for knowledge search, knowledge encoding, and knowledge incorporation. Finally, we propose how to improve existing systems based on theories from linguistics and cognitive science.

♻ ☆ Detecting and Mitigating Hallucinations in Machine Translation: Model Internal Workings Alone Do Well, Sentence Similarity Even Better

David Dale, Elena Voita, Loïc Barrault, Marta R. Costa-jussà

While the problem of hallucinations in neural machine translation has long been recognized, so far the progress on its alleviation is very little. Indeed, recently it turned out that without artificially encouraging models to hallucinate, previously existing methods fall short and even the standard sequence log-probability is more informative. It means that characteristics internal to the model can give much more information than we expect, and before using external models and measures, we first need to ask: how far can we go if we use nothing but the translation model itself ? We propose to use a method that evaluates the percentage of the source contribution to a generated translation. Intuitively, hallucinations are translations "detached" from the source, hence they can be identified by low source contribution. This method improves detection accuracy for the most severe hallucinations by a factor of 2 and is able to alleviate hallucinations at test time on par with the previous best approach that relies on external models. Next, if we move away from internal model characteristics and allow external tools, we show that using sentence similarity from cross-lingual embeddings further improves these results.

♻ ☆ GLUE-X: Evaluating Natural Language Understanding Models from an Out-of-distribution Generalization Perspective

Linyi Yang, Shuibai Zhang, Libo Qin, Yafu Li, Yidong Wang, Hanmeng Liu, Jindong Wang, Xing Xie, Yue Zhang

Pre-trained language models (PLMs) are known to improve the generalization performance of natural language understanding models by leveraging large amounts of data during the pre-training phase. However, the out-of-distribution (OOD) generalization problem remains a challenge in many NLP tasks, limiting the real-world deployment of these methods. This paper presents the first attempt at creating a unified benchmark named GLUE-X for evaluating OOD robustness in NLP models, highlighting the importance of OOD robustness and providing insights on how to measure the robustness of a model and how to improve it. The benchmark includes 13 publicly available datasets for OOD testing, and evaluations are conducted on 8 classic NLP tasks over 19 popularly used PLMs. Our findings confirm the need for improved OOD accuracy in NLP tasks, as significant performance degradation was observed in all settings compared to in-distribution (ID) accuracy.

comment: 17 pages, GLUE-X, OOD Generalization

♻ ☆ What do you MEME? Generating Explanations for Visual Semantic Role Labelling in Memes AAAI 2023

Shivam Sharma, Siddhant Agarwal, Tharun Suresh, Preslav Nakov, Md. Shad Akhtar, Tanmoy Chakraborty

Memes are powerful means for effective communication on social media. Their effortless amalgamation of viral visuals and compelling messages can have far-reaching implications with proper marketing. Previous research on memes has primarily focused on characterizing their affective spectrum and detecting whether the meme's message insinuates any intended harm, such as hate, offense, racism, etc. However, memes often use abstraction, which can be elusive. Here, we introduce a novel task - EXCLAIM, generating explanations for visual semantic role labeling in memes. To this end, we curate ExHVV, a novel dataset that offers natural language explanations of connotative roles for three types of entities - heroes, villains, and victims, encompassing 4,680 entities present in 3K memes. We also benchmark ExHVV with several strong unimodal and multimodal baselines. Moreover, we posit LUMEN, a novel multimodal, multi-task learning framework that endeavors to address EXCLAIM optimally by jointly learning to predict the correct semantic roles and correspondingly to generate suitable natural language explanations. LUMEN distinctly outperforms the best baseline across 18 standard natural language generation evaluation metrics. Our systematic evaluation and analyses demonstrate that characteristic multimodal cues required for adjudicating semantic roles are also helpful for generating suitable explanations.

comment: Accepted at AAAI 2023 (Main Track). 7 Pages (main content) + 2 Pages (Refs.); 3 Figures; 6 Tables; Paper ID: 10326 (AAAI'23)

♻ ☆ An Interdisciplinary Perspective on Evaluation and Experimental Design for Visual Text Analytics: Position Paper CCS

Kostiantyn Kucher, Nicole Sultanum, Angel Daza, Vasiliki Simaki, Maria Skeppstedt, Barbara Plank, Jean-Daniel Fekete, Narges Mahyar

Appropriate evaluation and experimental design are fundamental for empirical sciences, particularly in data-driven fields. Due to the successes in computational modeling of languages, for instance, research outcomes are having an increasingly immediate impact on end users. As the gap in adoption by end users decreases, the need increases to ensure that tools and models developed by the research communities and practitioners are reliable, trustworthy, and supportive of the users in their goals. In this position paper, we focus on the issues of evaluating visual text analytics approaches. We take an interdisciplinary perspective from the visualization and natural language processing communities, as we argue that the design and validation of visual text analytics include concerns beyond computational or visual/interactive methods on their own. We identify four key groups of challenges for evaluating visual text analytics approaches (data ambiguity, experimental design, user trust, and "big picture" concerns) and provide suggestions for research opportunities from an interdisciplinary perspective.

comment: Published in Proceedings of the 2022 IEEE Workshop on Evaluation and Beyond - Methodological Approaches to Visualization (BELIV '22). ACM 2012 CCS: Human-centered computing, Visualization, Visualization design and evaluation methods

♻ ☆ Diverse Demonstrations Improve In-context Compositional Generalization

Itay Levy, Ben Bogin, Jonathan Berant

In-context learning has shown great success in i.i.d semantic parsing splits, where the training and test sets are drawn from the same distribution. In this setup, models are typically prompted with demonstrations that are similar to the input question. However, in the setup of compositional generalization, where models are tested on outputs with structures that are absent from the training set, selecting similar demonstrations is insufficient, as often no example will be similar enough to the input. In this work, we propose a method to select diverse demonstrations that aims to collectively cover all of the structures required in the output program, in order to encourage the model to generalize to new structures from these demonstrations. We empirically show that combining diverse demonstrations with in-context learning substantially improves performance across three compositional generalization semantic parsing datasets in the pure in-context learning setup and when combined with finetuning.

♻ ☆ Analyzing Transformers in Embedding Space

Guy Dar, Mor Geva, Ankit Gupta, Jonathan Berant

Understanding Transformer-based models has attracted significant attention, as they lie at the heart of recent technological advances across machine learning. While most interpretability methods rely on running models over inputs, recent work has shown that a zero-pass approach, where parameters are interpreted directly without a forward/backward pass is feasible for some Transformer parameters, and for two-layer attention networks. In this work, we present a theoretical analysis where all parameters of a trained Transformer are interpreted by projecting them into the embedding space, that is, the space of vocabulary items they operate on. We derive a simple theoretical framework to support our arguments and provide ample evidence for its validity. First, an empirical analysis showing that parameters of both pretrained and fine-tuned models can be interpreted in embedding space. Second, we present two applications of our framework: (a) aligning the parameters of different models that share a vocabulary, and (b) constructing a classifier without training by ``translating'' the parameters of a fine-tuned classifier to parameters of a different model that was only pretrained. Overall, our findings open the door to interpretation methods that, at least in part, abstract away from model specifics and operate in the embedding space only.

♻ ☆ Interactive Model with Structural Loss for Language-based Abductive Reasoning

Linhao Li, Ming Xu, Yongfeng Dong, Xin Li, Ao Wang

The abductive natural language inference task ($\alpha$NLI) is proposed to infer the most plausible explanation between the cause and the event. In the $\alpha$NLI task, two observations are given, and the most plausible hypothesis is asked to pick out from the candidates. Existing methods model the relation between each candidate hypothesis separately and penalize the inference network uniformly. In this paper, we argue that it is unnecessary to distinguish the reasoning abilities among correct hypotheses; and similarly, all wrong hypotheses contribute the same when explaining the reasons of the observations. Therefore, we propose to group instead of ranking the hypotheses and design a structural loss called ``joint softmax focal loss'' in this paper. Based on the observation that the hypotheses are generally semantically related, we have designed a novel interactive language model aiming at exploiting the rich interaction among competing hypotheses. We name this new model for $\alpha$NLI: Interactive Model with Structural Loss (IMSL). The experimental results show that our IMSL has achieved the highest performance on the RoBERTa-large pretrained model, with ACC and AUC results increased by about 1\% and 5\% respectively.

comment: The paper is under consideration at Pattern Recognition Letters

♻ ☆ What Knowledge Is Needed? Towards Explainable Memory for kNN-MT Domain Adaptation

Wenhao Zhu, Shujian Huang, Yunzhe Lv, Xin Zheng, Jiajun Chen

kNN-MT presents a new paradigm for domain adaptation by building an external datastore, which usually saves all target language token occurrences in the parallel corpus. As a result, the constructed datastore is usually large and possibly redundant. In this paper, we investigate the interpretability issue of this approach: what knowledge does the NMT model need? We propose the notion of local correctness (LAC) as a new angle, which describes the potential translation correctness for a single entry and for a given neighborhood. Empirical study shows that our investigation successfully finds the conditions where the NMT model could easily fail and need related knowledge. Experiments on six diverse target domains and two language-pairs show that pruning according to local correctness brings a light and more explainable memory for kNN-MT domain adaptation.

♻ ☆ Technical Report -- Competition Solution for Prompt Tuning using Pretrained Language Model

Jiang-Long Song, Wu-He Zou, Feng Li, Xiao-Lei Qin, Wei-Dong Zhang

Prompt tuning recently becomes a hot-spot in the applications of large pretrained language models on specific downstream tasks. Regarding the Language Model as a Service (LMaaS), black-box tuning using derivative-free optimization (DFO) provides a novel approach to expand the practical scenarios of pretrained models and enrich the researches of few-shot learning. In this report, we present our solution in this competition that is based on the LMaaS scenario. Our solution consists of several modifications to BBTv2, including multiple label words, selection of P0, rolling update strategy, multi-task loss from MLP classifier, and finally using the ensemble method to further improve generalization ability. We also shared some strategies that we tried but didn't use in the final submission for further discussion. In the end we raised a question about the SNLI dataset and the impact on the results, as well as our concerns about the competition.

♻ ☆ Exploring Optimal Granularity for Extractive Summarization of Unstructured Health Records: Analysis of the Largest Multi-Institutional Archive of Health Records in Japan

Kenichiro Ando, Takashi Okumura, Mamoru Komachi, Hiromasa Horiguchi, Yuji Matsumoto

Automated summarization of clinical texts can reduce the burden of medical professionals. "Discharge summaries" are one promising application of the summarization, because they can be generated from daily inpatient records. Our preliminary experiment suggests that 20-31% of the descriptions in discharge summaries overlap with the content of the inpatient records. However, it remains unclear how the summaries should be generated from the unstructured source. To decompose the physician's summarization process, this study aimed to identify the optimal granularity in summarization. We first defined three types of summarization units with different granularities to compare the performance of the discharge summary generation: whole sentences, clinical segments, and clauses. We defined clinical segments in this study, aiming to express the smallest medically meaningful concepts. To obtain the clinical segments, it was necessary to automatically split the texts in the first stage of the pipeline. Accordingly, we compared rule-based methods and a machine learning method, and the latter outperformed the formers with an F1 score of 0.846 in the splitting task. Next, we experimentally measured the accuracy of extractive summarization using the three types of units, based on the ROUGE-1 metric, on a multi-institutional national archive of health records in Japan. The measured accuracies of extractive summarization using whole sentences, clinical segments, and clauses were 31.91, 36.15, and 25.18, respectively. We found that the clinical segments yielded higher accuracy than sentences and clauses. This result indicates that summarization of inpatient records demands finer granularity than sentence-oriented processing. Although we used only Japanese health records, it can be interpreted as follows: physicians extract "concepts of medical significance" from patient records and recombine them ...

♻ ☆ KRLS: Improving End-to-End Response Generation in Task Oriented Dialog with Reinforced Keywords Learning

Xiao Yu, Qingyang Wu, Kun Qian, Zhou Yu

In task-oriented dialogs, an informative and successful system response needs to include key information such as the phone number of a hotel. Therefore, we hypothesize that a model can achieve better overall performance by focusing on correctly generating key quantities. In this paper, we propose a new training algorithm, Keywords Reinforcement Learning with Next-word Sampling (KRLS), that utilizes Reinforcement Learning but avoids the time-consuming auto-regressive generation, and a fine-grained per-token reward function to help the model learn keywords generation more robustly. Empirical results show that the KRLS algorithm can achieve state-of-the-art performance on the inform, success, and combined score on the MultiWoZ benchmark dataset.

comment: added more explaination on the algorithm itself. result tables remain the same

♻ ☆ Deduplicating Training Data Mitigates Privacy Risks in Language Models ICML 2022

Nikhil Kandpal, Eric Wallace, Colin Raffel

Past work has shown that large language models are susceptible to privacy attacks, where adversaries generate sequences from a trained model and detect which sequences are memorized from the training set. In this work, we show that the success of these attacks is largely due to duplication in commonly used web-scraped training sets. We first show that the rate at which language models regenerate training sequences is superlinearly related to a sequence's count in the training set. For instance, a sequence that is present 10 times in the training data is on average generated ~1000 times more often than a sequence that is present only once. We next show that existing methods for detecting memorized sequences have near-chance accuracy on non-duplicated training sequences. Finally, we find that after applying methods to deduplicate training data, language models are considerably more secure against these types of privacy attacks. Taken together, our results motivate an increased focus on deduplication in privacy-sensitive applications and a reevaluation of the practicality of existing privacy attacks.

comment: ICML 2022 Camera Ready Version

♻ ☆ One Embedder, Any Task: Instruction-Finetuned Text Embeddings

Hongjin Su, Weijia Shi, Jungo Kasai, Yizhong Wang, Yushi Hu, Mari Ostendorf, Wen-tau Yih, Noah A. Smith, Luke Zettlemoyer, Tao Yu

We introduce INSTRUCTOR, a new method for computing text embeddings given task instructions: every text input is embedded together with instructions explaining the use case (e.g., task and domain descriptions). Unlike encoders from prior work that are more specialized, INSTRUCTOR is a single embedder that can generate text embeddings tailored to different downstream tasks and domains, without any further training. We first annotate instructions for 330 diverse tasks and train INSTRUCTOR on this multitask mixture with a contrastive loss. We evaluate INSTRUCTOR on 70 embedding evaluation tasks (66 of which are unseen during training), ranging from classification and information retrieval to semantic textual similarity and text generation evaluation. INSTRUCTOR, while having an order of magnitude fewer parameters than the previous best model, achieves state-of-the-art performance, with an average improvement of 3.4% compared to the previous best results on the 70 diverse datasets. Our analysis suggests that INSTRUCTOR is robust to changes in instructions, and that instruction finetuning mitigates the challenge of training a single model on diverse datasets. Our model, code, and data are available at https://instructor-embedding.github.io.

♻ ☆ Delving into the Openness of CLIP

Shuhuai Ren, Lei Li, Xuancheng Ren, Guangxiang Zhao, Xu Sun

Contrastive Language-Image Pre-training (CLIP) has demonstrated great potential in realizing open-vocabulary visual recognition in a matching style, due to its holistic use of natural language supervision that covers unconstrained real-world visual concepts. However, it is, in turn, also difficult to evaluate and analyze the openness of CLIP-like models, since they are in theory open to any vocabulary but the actual accuracy varies. To address the insufficiency of conventional studies on openness, we resort to an incremental perspective and define the extensibility, which essentially approximates the model's ability to deal with new visual concepts, by evaluating openness through vocabulary expansions. Our evaluation based on extensibility shows that CLIP-like models are hardly truly open and their performance degrades as the vocabulary expands to different degrees. Further analysis reveals that the over-estimation of openness is not because CLIP-like models fail to capture the general similarity of image and text features of novel visual concepts, but because of the confusion among competing text features, that is, they are not stable with respect to the vocabulary. In light of this, we propose to improve the openness of CLIP in the feature space by enforcing the distinguishability of text features. Our method retrieves relevant texts from the pre-training corpus to enhance prompts for inference, which boosts the extensibility and stability of CLIP even without fine-tuning.

comment: 22 pages, 12 figures. Code is available at https://github.com/lancopku/clip-openness

♻ ☆ NusaCrowd: Open Source Initiative for Indonesian NLP Resources

Samuel Cahyawijaya, Holy Lovenia, Alham Fikri Aji, Genta Indra Winata, Bryan Wilie, Rahmad Mahendra, Christian Wibisono, Ade Romadhony, Karissa Vincentio, Fajri Koto, Jennifer Santoso, David Moeljadi, Cahya Wirawan, Frederikus Hudi, Ivan Halim Parmonangan, Ika Alfina, Muhammad Satrio Wicaksono, Ilham Firdausi Putra, Samsul Rahmadani, Yulianti Oenang, Ali Akbar Septiandri, James Jaya, Kaustubh D. Dhole, Arie Ardiyanti Suryani, Rifki Afina Putri, Dan Su, Keith Stevens, Made Nindyatama Nityasya, Muhammad Farid Adilazuarda, Ryan Ignatius, Ryandito Diandaru, Tiezheng Yu, Vito Ghifari, Wenliang Dai, Yan Xu, Dyah Damapuspita, Cuk Tho, Ichwanul Muslim Karo Karo, Tirana Noor Fatyanosa, Ziwei Ji, Pascale Fung, Graham Neubig, Timothy Baldwin, Sebastian Ruder, Herry Sujaini, Sakriani Sakti, Ayu Purwarianti

We present NusaCrowd, a collaborative initiative to collect and unite existing resources for Indonesian languages, including opening access to previously non-public resources. Through this initiative, we have has brought together 137 datasets and 117 standardized data loaders. The quality of the datasets has been assessed manually and automatically, and their effectiveness has been demonstrated in multiple experiments. NusaCrowd's data collection enables the creation of the first zero-shot benchmarks for natural language understanding and generation in Indonesian and its local languages. Furthermore, NusaCrowd brings the creation of the first multilingual automatic speech recognition benchmark in Indonesian and its local languages. Our work is intended to help advance natural language processing research in under-represented languages.

♻ ☆ Unifying Vision, Text, and Layout for Universal Document Processing

Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal

We propose Universal Document Processing (UDOP), a foundation Document AI model which unifies text, image, and layout modalities together with varied task formats, including document understanding and generation. UDOP leverages the spatial correlation between textual content and document image to model image, text, and layout modalities with one uniform representation. With a novel Vision-Text-Layout Transformer, UDOP unifies pretraining and multi-domain downstream tasks into a prompt-based sequence generation scheme. UDOP is pretrained on both large-scale unlabeled document corpora using innovative self-supervised objectives and diverse labeled data. UDOP also learns to generate document images from text and layout modalities via masked image reconstruction. To the best of our knowledge, this is the first time in the field of document AI that one model simultaneously achieves high-quality neural document editing and content customization. Our method sets the state-of-the-art on 9 Document AI tasks, e.g., document understanding and QA, across diverse data domains like finance reports, academic papers, and websites. UDOP ranks first on the leaderboard of the Document Understanding Benchmark (DUE).

♻ ☆ Contextual-Lexicon Approach for Abusive Language Detection

Francielle Vargas, Fabiana Rodrigues de Góes, Isabelle Carvalho, Fabrício Benevenuto, Thiago Alexandre Salgueiro Pardo

Since a lexicon-based approach is more elegant scientifically, explaining the solution components and being easier to generalize to other applications, this paper provides a new approach for offensive language and hate speech detection on social media. Our approach embodies a lexicon of implicit and explicit offensive and swearing expressions annotated with contextual information. Due to the severity of the social media abusive comments in Brazil, and the lack of research in Portuguese, Brazilian Portuguese is the language used to validate the models. Nevertheless, our method may be applied to any other language. The conducted experiments show the effectiveness of the proposed approach, outperforming the current baseline methods for the Portuguese language.

comment: Please cite: https://aclanthology.org/2021.ranlp-1.161/

♻ ☆ Multi-Scales Data Augmentation Approach In Natural Language Inference For Artifacts Mitigation And Pre-Trained Model Optimization

Zhenyuan Lu

Machine learning models can reach high performance on benchmark natural language processing (NLP) datasets but fail in more challenging settings. We study this issue when a pre-trained model learns dataset artifacts in natural language inference (NLI), the topic of studying the logical relationship between a pair of text sequences. We provide a variety of techniques for analyzing and locating dataset artifacts inside the crowdsourced Stanford Natural Language Inference (SNLI) corpus. We study the stylistic pattern of dataset artifacts in the SNLI. To mitigate dataset artifacts, we employ a unique multi-scale data augmentation technique with two distinct frameworks: a behavioral testing checklist at the sentence level and lexical synonym criteria at the word level. Specifically, our combination method enhances our model's resistance to perturbation testing, enabling it to continuously outperform the pre-trained baseline.

♻ ☆ Can large language models reason about medical questions?

Valentin Liévin, Christoffer Egeberg Hother, Ole Winther

Although large language models (LLMs) often produce impressive outputs, it remains unclear how they perform in real-world scenarios requiring strong reasoning skills and expert domain knowledge. We set out to investigate whether GPT-3.5 (Codex and InstructGPT) can be applied to answer and reason about difficult real-world-based questions. We utilize two multiple-choice medical exam questions (USMLE and MedMCQA) and a medical reading comprehension dataset (PubMedQA). We investigate multiple prompting scenarios: Chain-of-Thought (CoT, think step-by-step), zero- and few-shot (prepending the question with question-answer exemplars) and retrieval augmentation (injecting Wikipedia passages into the prompt). For a subset of the USMLE questions, a medical expert reviewed and annotated the model's CoT. We found that InstructGPT can often read, reason and recall expert knowledge. Failure are primarily due to lack of knowledge and reasoning errors and trivial guessing heuristics are observed, e.g.\ too often predicting labels A and D on USMLE. Sampling and combining many completions overcome some of these limitations. Using 100 samples, Codex 5-shot CoT not only gives close to well-calibrated predictive probability but also achieves human-level performances on the three datasets. USMLE: 60.2%, MedMCQA: 57.5% and PubMedQA: 78.2%.

comment: 33 pages, 6 figures, to be submitted

☆ Character-Aware Models Improve Visual Text Rendering

Rosanne Liu, Dan Garrette, Chitwan Saharia, William Chan, Adam Roberts, Sharan Narang, Irina Blok, RJ Mical, Mohammad Norouzi, Noah Constant

Current image generation models struggle to reliably produce well-formed visual text. In this paper, we investigate a key contributing factor: popular text-to-image models lack character-level input features, making it much harder to predict a word's visual makeup as a series of glyphs. To quantify the extent of this effect, we conduct a series of controlled experiments comparing character-aware vs. character-blind text encoders. In the text-only domain, we find that character-aware models provide large gains on a novel spelling task (WikiSpell). Transferring these learnings onto the visual domain, we train a suite of image generation models, and show that character-aware variants outperform their character-blind counterparts across a range of novel text rendering tasks (our DrawText benchmark). Our models set a much higher state-of-the-art on visual spelling, with 30+ point accuracy gains over competitors on rare words, despite training on far fewer examples.

☆ Unleashing the Power of Visual Prompting At the Pixel Level

Junyang Wu, Xianhang Li, Chen Wei, Huiyu Wang, Alan Yuille, Yuyin Zhou, Cihang Xie

This paper presents a simple and effective visual prompting method for adapting pre-trained models to downstream recognition tasks. Our method includes two key designs. First, rather than directly adding together the prompt and the image, we treat the prompt as an extra and independent learnable component. We show that the strategy of reconciling the prompt and the image matters, and find that warping the prompt around a properly shrinked image empirically works the best. Second, we re-introduce two "old tricks" commonly used in building transferable adversarial examples, i.e., input diversity and gradient normalization, into visual prompting. These techniques improve optimization and enable the prompt to generalize better. We provide extensive experimental results to demonstrate the effectiveness of our method. Using a CLIP model, our prompting method sets a new record of 82.8% average accuracy across 12 popular classification datasets, substantially surpassing the prior art by +5.6%. It is worth noting that this prompting performance already outperforms linear probing by +2.1% and can even match fully fine-tuning in certain datasets. In addition, our prompting method shows competitive performance across different data scales and against distribution shifts. The code is publicly available at https://github.com/UCSC-VLAA/EVP.

☆ RangeAugment: Efficient Online Augmentation with Range Learning

Sachin Mehta, Saeid Naderiparizi, Fartash Faghri, Maxwell Horton, Lailin Chen, Ali Farhadi, Oncel Tuzel, Mohammad Rastegari

State-of-the-art automatic augmentation methods (e.g., AutoAugment and RandAugment) for visual recognition tasks diversify training data using a large set of augmentation operations. The range of magnitudes of many augmentation operations (e.g., brightness and contrast) is continuous. Therefore, to make search computationally tractable, these methods use fixed and manually-defined magnitude ranges for each operation, which may lead to sub-optimal policies. To answer the open question on the importance of magnitude ranges for each augmentation operation, we introduce RangeAugment that allows us to efficiently learn the range of magnitudes for individual as well as composite augmentation operations. RangeAugment uses an auxiliary loss based on image similarity as a measure to control the range of magnitudes of augmentation operations. As a result, RangeAugment has a single scalar parameter for search, image similarity, which we simply optimize via linear search. RangeAugment integrates seamlessly with any model and learns model- and task-specific augmentation policies. With extensive experiments on the ImageNet dataset across different networks, we show that RangeAugment achieves competitive performance to state-of-the-art automatic augmentation methods with 4-5 times fewer augmentation operations. Experimental results on semantic segmentation, object detection, foundation models, and knowledge distillation further shows RangeAugment's effectiveness.

comment: Technical report (22 pages including references and appendix)

☆ InstantAvatar: Learning Avatars from Monocular Video in 60 Seconds

Tianjian Jiang, Xu Chen, Jie Song, Otmar Hilliges

In this paper, we take a significant step towards real-world applicability of monocular neural avatar reconstruction by contributing InstantAvatar, a system that can reconstruct human avatars from a monocular video within seconds, and these avatars can be animated and rendered at an interactive rate. To achieve this efficiency we propose a carefully designed and engineered system, that leverages emerging acceleration structures for neural fields, in combination with an efficient empty space-skipping strategy for dynamic scenes. We also contribute an efficient implementation that we will make available for research purposes. Compared to existing methods, InstantAvatar converges 130x faster and can be trained in minutes instead of hours. It achieves comparable or even better reconstruction quality and novel pose synthesis results. When given the same time budget, our method significantly outperforms SoTA methods. InstantAvatar can yield acceptable visual quality in as little as 10 seconds training time.

comment: 12 pages

☆ Cross-modal Attention Congruence Regularization for Vision-Language Relation Alignment

Rohan Pandey, Rulin Shao, Paul Pu Liang, Ruslan Salakhutdinov, Louis-Philippe Morency

Despite recent progress towards scaling up multimodal vision-language models, these models are still known to struggle on compositional generalization benchmarks such as Winoground. We find that a critical component lacking from current vision-language models is relation-level alignment: the ability to match directional semantic relations in text (e.g., "mug in grass") with spatial relationships in the image (e.g., the position of the mug relative to the grass). To tackle this problem, we show that relation alignment can be enforced by encouraging the directed language attention from 'mug' to 'grass' (capturing the semantic relation 'in') to match the directed visual attention from the mug to the grass. Tokens and their corresponding objects are softly identified using the cross-modal attention. We prove that this notion of soft relation alignment is equivalent to enforcing congruence between vision and language attention matrices under a 'change of basis' provided by the cross-modal attention matrix. Intuitively, our approach projects visual attention into the language attention space to calculate its divergence from the actual language attention, and vice versa. We apply our Cross-modal Attention Congruence Regularization (CACR) loss to UNITER and improve on the state-of-the-art approach to Winoground.

☆ UNO-QA: An Unsupervised Anomaly-Aware Framework with Test-Time Clustering for OCTA Image Quality Assessment

Juntao Chen, Li Lin, Pujin Cheng, Yijin Huang, Xiaoying Tang

Medical image quality assessment (MIQA) is a vital prerequisite in various medical image analysis applications. Most existing MIQA algorithms are fully supervised that request a large amount of annotated data. However, annotating medical images is time-consuming and labor-intensive. In this paper, we propose an unsupervised anomaly-aware framework with test-time clustering for optical coherence tomography angiography (OCTA) image quality assessment in a setting wherein only a set of high-quality samples are accessible in the training phase. Specifically, a feature-embedding-based low-quality representation module is proposed to quantify the quality of OCTA images and then to discriminate between outstanding quality and non-outstanding quality. Within the non-outstanding quality class, to further distinguish gradable images from ungradable ones, we perform dimension reduction and clustering of multi-scale image features extracted by the trained OCTA quality representation network. Extensive experiments are conducted on one publicly accessible dataset sOCTA-3*3-10k, with superiority of our proposed framework being successfully established.

comment: submitted to ISBI2023

☆ Does CLIP Bind Concepts? Probing Compositionality in Large Image Models

Martha Lewis, Qinan Yu, Jack Merullo, Ellie Pavlick

Large-scale models combining text and images have made incredible progress in recent years. However, they can still fail at tasks requiring compositional knowledge, such as correctly picking out a red cube from a picture of multiple shapes. We examine the ability of CLIP (Radford et al., 2021), to caption images requiring compositional knowledge. We implement five compositional language models to probe the kinds of structure that CLIP may be using, and develop a novel training algorithm, Compositional Skipgram for Images (CoSI), to train these models. We look at performance in attribute-based tasks, requiring the identification of a particular combination of attribute and object (such as "red cube"), and in relational settings, where the spatial relation between two shapes (such as "cube behind sphere") must be identified. We find that in some conditions, CLIP is able to learn attribute-object labellings, and to generalize to unseen attribute-object combinations. However, we also see evidence that CLIP is not able to bind features together reliably. Moreover, CLIP is not able to reliably learn relations between objects, whereas some compositional models are able to learn these perfectly. Of the five models we developed, none were able to generalize to unseen relations.

☆ A Survey of Deep Learning for Mathematical Reasoning

Pan Lu, Liang Qiu, Wenhao Yu, Sean Welleck, Kai-Wei Chang

Mathematical reasoning is a fundamental aspect of human intelligence and is applicable in various fields, including science, engineering, finance, and everyday life. The development of artificial intelligence (AI) systems capable of solving math problems and proving theorems has garnered significant interest in the fields of machine learning and natural language processing. For example, mathematics serves as a testbed for aspects of reasoning that are challenging for powerful deep learning models, driving new algorithmic and modeling advances. On the other hand, recent advances in large-scale neural language models have opened up new benchmarks and opportunities to use deep learning for mathematical reasoning. In this survey paper, we review the key tasks, datasets, and methods at the intersection of mathematical reasoning and deep learning over the past decade. We also evaluate existing benchmarks and methods, and discuss future research directions in this domain.

comment: 24 pages, 2 figures, 8 tables. The repository is available at https://github.com/lupantech/dl4math

☆ DePlot: One-shot visual language reasoning by plot-to-table translation

Fangyu Liu, Julian Martin Eisenschlos, Francesco Piccinno, Syrine Krichene, Chenxi Pang, Kenton Lee, Mandar Joshi, Wenhu Chen, Nigel Collier, Yasemin Altun

Visual language such as charts and plots is ubiquitous in the human world. Comprehending plots and charts requires strong reasoning skills. Prior state-of-the-art (SOTA) models require at least tens of thousands of training examples and their reasoning capabilities are still much limited, especially on complex human-written queries. This paper presents the first one-shot solution to visual language reasoning. We decompose the challenge of visual language reasoning into two steps: (1) plot-to-text translation, and (2) reasoning over the translated text. The key in this method is a modality conversion module, named as DePlot, which translates the image of a plot or chart to a linearized table. The output of DePlot can then be directly used to prompt a pretrained large language model (LLM), exploiting the few-shot reasoning capabilities of LLMs. To obtain DePlot, we standardize the plot-to-table task by establishing unified task formats and metrics, and train DePlot end-to-end on this task. DePlot can then be used off-the-shelf together with LLMs in a plug-and-play fashion. Compared with a SOTA model finetuned on more than >28k data points, DePlot+LLM with just one-shot prompting achieves a 24.0% improvement over finetuned SOTA on human-written queries from the task of chart QA.

☆ Recycling diverse models for out-of-distribution generalization

Alexandre Ramé, Kartik Ahuja, Jianyu Zhang, Matthieu Cord, Léon Bottou, David Lopez-Paz

Foundation models are redefining how AI systems are built. Practitioners now follow a standard procedure to build their machine learning solutions: download a copy of a foundation model, and fine-tune it using some in-house data about the target task of interest. Consequently, the Internet is swarmed by a handful of foundation models fine-tuned on many diverse tasks. Yet, these individual fine-tunings often lack strong generalization and exist in isolation without benefiting from each other. In our opinion, this is a missed opportunity, as these specialized models contain diverse features. Based on this insight, we propose model recycling, a simple strategy that leverages multiple fine-tunings of the same foundation model on diverse auxiliary tasks, and repurposes them as rich and diverse initializations for the target task. Specifically, model recycling fine-tunes in parallel each specialized model on the target task, and then averages the weights of all target fine-tunings into a final model. Empirically, we show that model recycling maximizes model diversity by benefiting from diverse auxiliary tasks, and achieves a new state of the art on the reference DomainBed benchmark for out-of-distribution generalization. Looking forward, model recycling is a contribution to the emerging paradigm of updatable machine learning where, akin to open-source software development, the community collaborates to incrementally and reliably update machine learning models.

comment: 24 pages, 7 tables, 13 figures

☆ QuantArt: Quantizing Image Style Transfer Towards High Visual Fidelity

Siyu Huang, Jie An, Donglai Wei, Jiebo Luo, Hanspeter Pfister

The mechanism of existing style transfer algorithms is by minimizing a hybrid loss function to push the generated image toward high similarities in both content and style. However, this type of approach cannot guarantee visual fidelity, i.e., the generated artworks should be indistinguishable from real ones. In this paper, we devise a new style transfer framework called QuantArt for high visual-fidelity stylization. QuantArt pushes the latent representation of the generated artwork toward the centroids of the real artwork distribution with vector quantization. By fusing the quantized and continuous latent representations, QuantArt allows flexible control over the generated artworks in terms of content preservation, style similarity, and visual fidelity. Experiments on various style transfer settings show that our QuantArt framework achieves significantly higher visual fidelity compared with the existing style transfer methods.

☆ HouseCat6D -- A Large-Scale Multi-Modal Category Level 6D Object Pose Dataset with Household Objects in Realistic Scenarios

HyunJun Jung, Shun-Cheng Wu, Patrick Ruhkamp, Hannah Schieber, Pengyuan Wang, Giulia Rizzoli, Hongcheng Zhao, Sven Damian Meier, Daniel Roth, Nassir Navab, Benjamin Busam

Estimating the 6D pose of objects is one of the major fields in 3D computer vision. Since the promising outcomes from instance-level pose estimation, the research trends are heading towards category-level pose estimation for more practical application scenarios. However, unlike well-established instance-level pose datasets, available category-level datasets lack annotation quality and provided pose quantity. We propose the new category level 6D pose dataset HouseCat6D featuring 1) Multi-modality of Polarimetric RGB+P and Depth, 2) Highly diverse 194 objects of 10 household object categories including 2 photometrically challenging categories, 3) High-quality pose annotation with an error range of only 1.35 mm to 1.74 mm, 4) 41 large scale scenes with extensive viewpoint coverage, 5) Checkerboard-free environment throughout the entire scene. We also provide benchmark results of state-of-the-art category-level pose estimation networks.

☆ Scene Change Detection Using Multiscale Cascade Residual Convolutional Neural Networks

Daniel F. S. Santos, Rafael G. Pires, Danilo Colombo, João P. Papa

Scene change detection is an image processing problem related to partitioning pixels of a digital image into foreground and background regions. Mostly, visual knowledge-based computer intelligent systems, like traffic monitoring, video surveillance, and anomaly detection, need to use change detection techniques. Amongst the most prominent detection methods, there are the learning-based ones, which besides sharing similar training and testing protocols, differ from each other in terms of their architecture design strategies. Such architecture design directly impacts on the quality of the detection results, and also in the device resources capacity, like memory. In this work, we propose a novel Multiscale Cascade Residual Convolutional Neural Network that integrates multiscale processing strategy through a Residual Processing Module, with a Segmentation Convolutional Neural Network. Experiments conducted on two different datasets support the effectiveness of the proposed approach, achieving average overall $\boldsymbol{F\text{-}measure}$ results of $\boldsymbol{0.9622}$ and $\boldsymbol{0.9664}$ over Change Detection 2014 and PetrobrasROUTES datasets respectively, besides comprising approximately eight times fewer parameters. Such obtained results place the proposed technique amongst the top four state-of-the-art scene change detection methods.

comment: Published in: 2020 33rd SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI)

☆ DDIPNet and DDIPNet+: Discriminant Deep Image Prior Networks for Remote Sensing Image Classification

Daniel F. S. Santos, Rafael G. Pires, Leandro A. Passos, João P. Papa

Research on remote sensing image classification significantly impacts essential human routine tasks such as urban planning and agriculture. Nowadays, the rapid advance in technology and the availability of many high-quality remote sensing images create a demand for reliable automation methods. The current paper proposes two novel deep learning-based architectures for image classification purposes, i.e., the Discriminant Deep Image Prior Network and the Discriminant Deep Image Prior Network+, which combine Deep Image Prior and Triplet Networks learning strategies. Experiments conducted over three well-known public remote sensing image datasets achieved state-of-the-art results, evidencing the effectiveness of using deep image priors for remote sensing image classification.

comment: Published in: 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS

☆ ADAS: A Simple Active-and-Adaptive Baseline for Cross-Domain 3D Semantic Segmentation

Ben Fei, Siyuan Huang, Jiakang Yuan, Botian Shi, Bo Zhang, Tao Chen, Min Dou, Yu Qiao

State-of-the-art 3D semantic segmentation models are trained on the off-the-shelf public benchmarks, but they often face the major challenge when these well-trained models are deployed to a new domain. In this paper, we propose an Active-and-Adaptive Segmentation (ADAS) baseline to enhance the weak cross-domain generalization ability of a well-trained 3D segmentation model, and bridge the point distribution gap between domains. Specifically, before the cross-domain adaptation stage begins, ADAS performs an active sampling operation to select a maximally-informative subset from both source and target domains for effective adaptation, reducing the adaptation difficulty under 3D scenarios. Benefiting from the rise of multi-modal 2D-3D datasets, ADAS utilizes a cross-modal attention-based feature fusion module that can extract a representative pair of image features and point features to achieve a bi-directional image-point feature interaction for better safe adaptation. Experimentally, ADAS is verified to be effective in many cross-domain settings including: 1) Unsupervised Domain Adaptation (UDA), which means that all samples from target domain are unlabeled; 2) Unsupervised Few-shot Domain Adaptation (UFDA) which means that only a few unlabeled samples are available in the unlabeled target domain; 3) Active Domain Adaptation (ADA) which means that the selected target samples by ADAS are manually annotated. Their results demonstrate that ADAS achieves a significant accuracy gain by easily coupling ADAS with self-training methods or off-the-shelf UDA works.

☆ Masked Event Modeling: Self-Supervised Pretraining for Event Cameras

Simon Klenk, David Bonello, Lukas Koestler, Daniel Cremers

Event cameras offer the capacity to asynchronously capture brightness changes with low latency, high temporal resolution, and high dynamic range. Deploying deep learning methods for classification or other tasks to these sensors typically requires large labeled datasets. Since the amount of labeled event data is tiny compared to the bulk of labeled RGB imagery, the progress of event-based vision has remained limited. To reduce the dependency on labeled event data, we introduce Masked Event Modeling (MEM), a self-supervised pretraining framework for events. Our method pretrains a neural network on unlabeled events, which can originate from any event camera recording. Subsequently, the pretrained model is finetuned on a downstream task leading to an overall better performance while requiring fewer labels. Our method outperforms the state-of-the-art on N-ImageNet, N-Cars, and N-Caltech101, increasing the object classification accuracy on N-ImageNet by 7.96%. We demonstrate that Masked Event Modeling is superior to RGB-based pretraining on a real world dataset.

☆ Modeling Human Eye Movements with Neural Networks in a Maze-Solving Task

Jason Li, Nicholas Watters, Yingting, Wang, Hansem Sohn, Mehrdad Jazayeri

From smoothly pursuing moving objects to rapidly shifting gazes during visual search, humans employ a wide variety of eye movement strategies in different contexts. While eye movements provide a rich window into mental processes, building generative models of eye movements is notoriously difficult, and to date the computational objectives guiding eye movements remain largely a mystery. In this work, we tackled these problems in the context of a canonical spatial planning task, maze-solving. We collected eye movement data from human subjects and built deep generative models of eye movements using a novel differentiable architecture for gaze fixations and gaze shifts. We found that human eye movements are best predicted by a model that is optimized not to perform the task as efficiently as possible but instead to run an internal simulation of an object traversing the maze. This not only provides a generative model of eye movements in this task but also suggests a computational theory for how humans solve the task, namely that humans use mental simulation.

☆ Weakly supervised training of universal visual concepts for multi-domain semantic segmentation

Petra Bevandić, Marin Oršić, Ivan Grubišić, Josip Šarić, Siniša Šegvić

Deep supervised models have an unprecedented capacity to absorb large quantities of training data. Hence, training on multiple datasets becomes a method of choice towards strong generalization in usual scenes and graceful performance degradation in edge cases. Unfortunately, different datasets often have incompatible labels. For instance, the Cityscapes road class subsumes all driving surfaces, while Vistas defines separate classes for road markings, manholes etc. Furthermore, many datasets have overlapping labels. For instance, pickups are labeled as trucks in VIPER, cars in Vistas, and vans in ADE20k. We address this challenge by considering labels as unions of universal visual concepts. This allows seamless and principled learning on multi-domain dataset collections without requiring any relabeling effort. Our method achieves competitive within-dataset and cross-dataset generalization, as well as ability to learn visual concepts which are not separately labeled in any of the training datasets. Experiments reveal competitive or state-of-the-art performance on two multi-domain dataset collections and on the WildDash 2 benchmark.

comment: 23 pages, 10 figures, 9 tables

☆ Image quality prediction using synthetic and natural codebooks: comparative results

Maxim Koroteev, Kirill Aistov, Valeriy Berezovskiy, Pavel Frolov

We investigate a model for image/video quality assessment based on building a set of codevectors representing in a sense some basic properties of images, similar to well-known CORNIA model. We analyze the codebook building method and propose some modifications for it. Also the algorithm is investigated from the point of inference time reduction. Both natural and synthetic images are used for building codebooks and some analysis of synthetic images used for codebooks is provided. It is demonstrated the results on quality assessment may be improves with the use if synthetic images for codebook construction. We also demonstrate regimes of the algorithm in which real time execution on CPU is possible for sufficiently high correlations with mean opinion score (MOS). Various pooling strategies are considered as well as the problem of metric sensitivity to bitrate.

comment: 18 pages, 8 figures

☆ Which Pixel to Annotate: a Label-Efficient Nuclei Segmentation Framework

Wei Lou, Haofeng Li, Guanbin Li, Xiaoguang Han, Xiang Wan

Recently deep neural networks, which require a large amount of annotated samples, have been widely applied in nuclei instance segmentation of H\&E stained pathology images. However, it is inefficient and unnecessary to label all pixels for a dataset of nuclei images which usually contain similar and redundant patterns. Although unsupervised and semi-supervised learning methods have been studied for nuclei segmentation, very few works have delved into the selective labeling of samples to reduce the workload of annotation. Thus, in this paper, we propose a novel full nuclei segmentation framework that chooses only a few image patches to be annotated, augments the training set from the selected samples, and achieves nuclei segmentation in a semi-supervised manner. In the proposed framework, we first develop a novel consistency-based patch selection method to determine which image patches are the most beneficial to the training. Then we introduce a conditional single-image GAN with a component-wise discriminator, to synthesize more training samples. Lastly, our proposed framework trains an existing segmentation model with the above augmented samples. The experimental results show that our proposed method could obtain the same-level performance as a fully-supervised baseline by annotating less than 5% pixels on some benchmarks.

comment: IEEE TMI 2022, Released code: https://github.com/lhaof/NuSeg

☆ Towards Unsupervised Visual Reasoning: Do Off-The-Shelf Features Know How to Reason?

Monika Wysoczańska, Tom Monnier, Tomasz Trzciński, David Picard

Recent advances in visual representation learning allowed to build an abundance of powerful off-the-shelf features that are ready-to-use for numerous downstream tasks. This work aims to assess how well these features preserve information about the objects, such as their spatial location, their visual properties and their relative relationships. We propose to do so by evaluating them in the context of visual reasoning, where multiple objects with complex relationships and different attributes are at play. More specifically, we introduce a protocol to evaluate visual representations for the task of Visual Question Answering. In order to decouple visual feature extraction from reasoning, we design a specific attention-based reasoning module which is trained on the frozen visual representations to be evaluated, in a spirit similar to standard feature evaluations relying on shallow networks. We compare two types of visual representations, densely extracted local features and object-centric ones, against the performances of a perfect image representation using ground truth. Our main findings are two-fold. First, despite excellent performances on classical proxy tasks, such representations fall short for solving complex reasoning problem. Second, object-centric features better preserve the critical information necessary to perform visual reasoning. In our proposed framework we show how to methodologically approach this evaluation.

☆ Quantifying and Visualizing Vascular Branching Geometry with Micro-CT: Normalization of Intra- and Inter-Specimen Variations

Timothy L. Kline

Micro-CT images of the renal arteries of intact rat kidneys, which had their vasculature injected with the contrast agent polymer Microfil, were characterized. Measurement of inter-branch segment properties and the hierarchical structure of the vessel trees were computed by an automated algorithmic approach. The perfusion territories of the different kidneys, as well as the local diameters of the segmented vasculature were mapped onto the representative structures and visually explored. Various parameters were compared in order to outline key geometrical properties, properties which were shown to not have a wide range of inter-specimen variation. It is shown that the fractal scaling in non-symmetric branching reveals itself differently, than in symmetric branching (e.g., in the lung the mean bronchial diameters at each generation are closely related). Also, perfused tissue is shown to have very little inter-specimen variation and therefore could be used in future studies related to characterizing various disease states of tissues and organs based on vascular branching geometry.

☆ High-resolution canopy height map in the Landes forest (France) based on GEDI, Sentinel-1, and Sentinel-2 data with a deep learning approach

Martin Schwartz, Philippe Ciais, Catherine Ottlé, Aurelien De Truchis, Cedric Vega, Ibrahim Fayad, Martin Brandt, Rasmus Fensholt, Nicolas Baghdadi, François Morneau, David Morin, Dominique Guyon, Sylvia Dayau, Jean-Pierre Wigneron

In intensively managed forests in Europe, where forests are divided into stands of small size and may show heterogeneity within stands, a high spatial resolution (10 - 20 meters) is arguably needed to capture the differences in canopy height. In this work, we developed a deep learning model based on multi-stream remote sensing measurements to create a high-resolution canopy height map over the "Landes de Gascogne" forest in France, a large maritime pine plantation of 13,000 km$^2$ with flat terrain and intensive management. This area is characterized by even-aged and mono-specific stands, of a typical length of a few hundred meters, harvested every 35 to 50 years. Our deep learning U-Net model uses multi-band images from Sentinel-1 and Sentinel-2 with composite time averages as input to predict tree height derived from GEDI waveforms. The evaluation is performed with external validation data from forest inventory plots and a stereo 3D reconstruction model based on Skysat imagery available at specific locations. We trained seven different U-net models based on a combination of Sentinel-1 and Sentinel-2 bands to evaluate the importance of each instrument in the dominant height retrieval. The model outputs allow us to generate a 10 m resolution canopy height map of the whole "Landes de Gascogne" forest area for 2020 with a mean absolute error of 2.02 m on the Test dataset. The best predictions were obtained using all available satellite layers from Sentinel-1 and Sentinel-2 but using only one satellite source also provided good predictions. For all validation datasets in coniferous forests, our model showed better metrics than previous canopy height models available in the same region.

comment: 39 pages, 16 figures + supplementary contents

☆ Eff-3DPSeg: 3D organ-level plant shoot segmentation using annotation-efficient point clouds

Liyi Luo, Xintong Jiang, Yu Yang, Eugene Roy Antony Samy, Mark Lefsrud, Valerio Hoyos-Villegas, Shangpeng Sun

Reliable and automated 3D plant shoot segmentation is a core prerequisite for the extraction of plant phenotypic traits at the organ level. Combining deep learning and point clouds can provide effective ways to address the challenge. However, fully supervised deep learning methods require datasets to be point-wise annotated, which is extremely expensive and time-consuming. In our work, we proposed a novel weakly supervised framework, Eff-3DPSeg, for 3D plant shoot segmentation. First, high-resolution point clouds of soybean were reconstructed using a low-cost photogrammetry system, and the Meshlab-based Plant Annotator was developed for plant point cloud annotation. Second, a weakly-supervised deep learning method was proposed for plant organ segmentation. The method contained: (1) Pretraining a self-supervised network using Viewpoint Bottleneck loss to learn meaningful intrinsic structure representation from the raw point clouds; (2) Fine-tuning the pre-trained model with about only 0.5% points being annotated to implement plant organ segmentation. After, three phenotypic traits (stem diameter, leaf width, and leaf length) were extracted. To test the generality of the proposed method, the public dataset Pheno4D was included in this study. Experimental results showed that the weakly-supervised network obtained similar segmentation performance compared with the fully-supervised setting. Our method achieved 95.1%, 96.6%, 95.8% and 92.2% in the Precision, Recall, F1-score, and mIoU for stem leaf segmentation and 53%, 62.8% and 70.3% in the AP, AP@25, and AP@50 for leaf instance segmentation. This study provides an effective way for characterizing 3D plant architecture, which will become useful for plant breeders to enhance selection processes.

comment: This paper has been submitted to Plant Phenomics

☆ Self-Pair: Synthesizing Changes from Single Source for Object Change Detection in Remote Sensing Imagery WACV2023

Minseok Seo, Hakjin Lee, Yongjin Jeon, Junghoon Seo

For change detection in remote sensing, constructing a training dataset for deep learning models is difficult due to the requirements of bi-temporal supervision. To overcome this issue, single-temporal supervision which treats change labels as the difference of two semantic masks has been proposed. This novel method trains a change detector using two spatially unrelated images with corresponding semantic labels such as building. However, training on unpaired datasets could confuse the change detector in the case of pixels that are labeled unchanged but are visually significantly different. In order to maintain the visual similarity in unchanged area, in this paper, we emphasize that the change originates from the source image and show that manipulating the source image as an after-image is crucial to the performance of change detection. Extensive experiments demonstrate the importance of maintaining visual information between pre- and post-event images, and our method outperforms existing methods based on single-temporal supervision. code is available at https://github.com/seominseok0429/Self-Pair-for-Change-Detection.

comment: This paper has been accepted by WACV2023

☆ A Comprehensive Study and Comparison of the Robustness of 3D Object Detectors Against Adversarial Attacks

Yifan Zhang, Junhui Hou, Yixuan Yuan

Deep learning-based 3D object detectors have made significant progress in recent years and have been deployed in a wide range of applications. It is crucial to understand the robustness of detectors against adversarial attacks when employing detectors in security-critical applications. In this paper, we make the first attempt to conduct a thorough evaluation and analysis of the robustness of 3D detectors under adversarial attacks. Specifically, we first extend three kinds of adversarial attacks to the 3D object detection task to benchmark the robustness of state-of-the-art 3D object detectors against attacks on KITTI and Waymo datasets, subsequently followed by the analysis of the relationship between robustness and properties of detectors. Then, we explore the transferability of cross-model, cross-task, and cross-data attacks. We finally conduct comprehensive experiments of defense for 3D detectors, demonstrating that simple transformations like flipping are of little help in improving robustness when the strategy of transformation imposed on input point cloud data is exposed to attackers. Our findings will facilitate investigations in understanding and defending the adversarial attacks against 3D object detectors to advance this field.

☆ StyleDomain: Analysis of StyleSpace for Domain Adaptation of StyleGAN

Aibek Alanov, Vadim Titov, Maksim Nakhodnov, Dmitry Vetrov

Domain adaptation of GANs is a problem of fine-tuning the state-of-the-art GAN models (e.g. StyleGAN) pretrained on a large dataset to a specific domain with few samples (e.g. painting faces, sketches, etc.). While there are a great number of methods that tackle this problem in different ways there are still many important questions that remain unanswered. In this paper, we provide a systematic and in-depth analysis of the domain adaptation problem of GANs, focusing on the StyleGAN model. First, we perform a detailed exploration of the most important parts of StyleGAN that are responsible for adapting the generator to a new domain depending on the similarity between the source and target domains. In particular, we show that affine layers of StyleGAN can be sufficient for fine-tuning to similar domains. Second, inspired by these findings, we investigate StyleSpace to utilize it for domain adaptation. We show that there exist directions in the StyleSpace that can adapt StyleGAN to new domains. Further, we examine these directions and discover their many surprising properties. Finally, we leverage our analysis and findings to deliver practical improvements and applications in such standard tasks as image-to-image translation and cross-domain morphing.

comment: Preprint

☆ CSMPQ:Class Separability Based Mixed-Precision Quantization

Mingkai Wang, Taisong Jin, Miaohui Zhang, Zhengtao Yu

Mixed-precision quantization has received increasing attention for its capability of reducing the computational burden and speeding up the inference time. Existing methods usually focus on the sensitivity of different network layers, which requires a time-consuming search or training process. To this end, a novel mixed-precision quantization method, termed CSMPQ, is proposed. Specifically, the TF-IDF metric that is widely used in natural language processing (NLP) is introduced to measure the class separability of layer-wise feature maps. Furthermore, a linear programming problem is designed to derive the optimal bit configuration for each layer. Without any iterative process, the proposed CSMPQ achieves better compression trade-offs than the state-of-the-art quantization methods. Specifically, CSMPQ achieves 73.03$\%$ Top-1 acc on ResNet-18 with only 59G BOPs for QAT, and 71.30$\%$ top-1 acc with only 1.5Mb on MobileNetV2 for PTQ.

☆ ParallelNet: Multi-mode Trajectory Prediction by Multi-mode Trajectory Fusion

Fei Wu, Luoyu Chen, Hao Lu

Level 5 Autonomous Driving, a technology that a fully automated vehicle (AV) requires no human intervention, has raised serious concerns on safety and stability before widespread use. The capability of understanding and predicting future motion trajectory of road objects can help AV plan a path that is safe and easy to control. In this paper, we propose a network architecture that parallelizes multiple convolutional neural network backbones and fuses features to make multi-mode trajectory prediction. In the 2020 ICRA Nuscene Prediction challenge, our model ranks 15th on the leaderboard across all teams.

comment: 8 pages,8 figures

☆ Redistribution of Weights and Activations for AdderNet Quantization

Ying Nie, Kai Han, Haikang Diao, Chuanjian Liu, Enhua Wu, Yunhe Wang

Adder Neural Network (AdderNet) provides a new way for developing energy-efficient neural networks by replacing the expensive multiplications in convolution with cheaper additions (i.e.l1-norm). To achieve higher hardware efficiency, it is necessary to further study the low-bit quantization of AdderNet. Due to the limitation that the commutative law in multiplication does not hold in l1-norm, the well-established quantization methods on convolutional networks cannot be applied on AdderNets. Thus, the existing AdderNet quantization techniques propose to use only one shared scale to quantize both the weights and activations simultaneously. Admittedly, such an approach can keep the commutative law in the l1-norm quantization process, while the accuracy drop after low-bit quantization cannot be ignored. To this end, we first thoroughly analyze the difference on distributions of weights and activations in AdderNet and then propose a new quantization algorithm by redistributing the weights and the activations. Specifically, the pre-trained full-precision weights in different kernels are clustered into different groups, then the intra-group sharing and inter-group independent scales can be adopted. To further compensate the accuracy drop caused by the distribution difference, we then develop a lossless range clamp scheme for weights and a simple yet effective outliers clamp strategy for activations. Thus, the functionality of full-precision weights and the representation ability of full-precision activations can be fully preserved. The effectiveness of the proposed quantization method for AdderNet is well verified on several benchmarks, e.g., our 4-bit post-training quantized adder ResNet-18 achieves an 66.5% top-1 accuracy on the ImageNet with comparable energy efficiency, which is about 8.5% higher than that of the previous AdderNet quantization methods.

☆ CGCV:Context Guided Correlation Volume for Optical Flow Neural Networks

Jiangpeng Li, Yan Niu

Optical flow, which computes the apparent motion from a pair of video frames, is a critical tool for scene motion estimation. Correlation volume is the central component of optical flow computational neural models. It estimates the pairwise matching costs between cross-frame features, and is then used to decode optical flow. However, traditional correlation volume is frequently noisy, outlier-prone, and sensitive to motion blur. We observe that, although the recent RAFT algorithm also adopts the traditional correlation volume, its additional context encoder provides semantically representative features to the flow decoder, implicitly compensating for the deficiency of the correlation volume. However, the benefits of this context encoder has been barely discussed or exploited. In this paper, we first investigate the functionality of RAFT's context encoder, then propose a new Context Guided Correlation Volume (CGCV) via gating and lifting schemes. CGCV can be universally integrated with RAFT-based flow computation methods for enhanced performance, especially effective in the presence of motion blur, de-focus blur and atmospheric effects. By incorporating the proposed CGCV with previous Global Motion Aggregation (GMA) method, at a minor cost of 0.5% extra parameters, the rank of GMA is lifted by 23 places on KITTI 2015 Leader Board, and 3 places on Sintel Leader Board. Moreover, at a similar model size, our correlation volume achieves competitive or superior performance to state of the art peer supervised models that employ Transformers or Graph Reasoning, as verified by extensive experiments.

☆ Hoyer regularizer is all you need for ultra low-latency spiking neural networks

Gourav Datta, Zeyu Liu, Peter A. Beerel

Spiking Neural networks (SNN) have emerged as an attractive spatio-temporal computing paradigm for a wide range of low-power vision tasks. However, state-of-the-art (SOTA) SNN models either incur multiple time steps which hinder their deployment in real-time use cases or increase the training complexity significantly. To mitigate this concern, we present a training framework (from scratch) for one-time-step SNNs that uses a novel variant of the recently proposed Hoyer regularizer. We estimate the threshold of each SNN layer as the Hoyer extremum of a clipped version of its activation map, where the clipping threshold is trained using gradient descent with our Hoyer regularizer. This approach not only downscales the value of the trainable threshold, thereby emitting a large number of spikes for weight update with a limited number of iterations (due to only one time step) but also shifts the membrane potential values away from the threshold, thereby mitigating the effect of noise that can degrade the SNN accuracy. Our approach outperforms existing spiking, binary, and adder neural networks in terms of the accuracy-FLOPs trade-off for complex image recognition tasks. Downstream experiments on object detection also demonstrate the efficacy of our approach.

comment: 16 pages, 4 figures

☆ Seafloor-Invariant Caustics Removal from Underwater Imagery

Panagiotis Agrafiotis, Konstantinos Karantzalos, Andreas Georgopoulos

Mapping the seafloor with underwater imaging cameras is of significant importance for various applications including marine engineering, geology, geomorphology, archaeology and biology. For shallow waters, among the underwater imaging challenges, caustics i.e., the complex physical phenomena resulting from the projection of light rays being refracted by the wavy surface, is likely the most crucial one. Caustics is the main factor during underwater imaging campaigns that massively degrade image quality and affect severely any 2D mosaicking or 3D reconstruction of the seabed. In this work, we propose a novel method for correcting the radiometric effects of caustics on shallow underwater imagery. Contrary to the state-of-the-art, the developed method can handle seabed and riverbed of any anaglyph, correcting the images using real pixel information, thus, improving image matching and 3D reconstruction processes. In particular, the developed method employs deep learning architectures in order to classify image pixels to "non-caustics" and "caustics". Then, exploits the 3D geometry of the scene to achieve a pixel-wise correction, by transferring appropriate color values between the overlapping underwater images. Moreover, to fill the current gap, we have collected, annotated and structured a real-world caustic dataset, namely R-CAUSTIC, which is openly available. Overall, based on the experimental results and validation the developed methodology is quite promising in both detecting caustics and reconstructing their intensity.

comment: Submitted to the IEEE Journal of Oceanic Engineering (IEEE-JOE), under review as of December 2022

☆ Goal-oriented Autonomous Driving

Yihan Hu, Jiazhi Yang, Li Chen, Keyu Li, Chonghao Sima, Xizhou Zhu, Siqi Chai, Senyao Du, Tianwei Lin, Wenhai Wang, Lewei Lu, Xiaosong Jia, Qiang Liu, Jifeng Dai, Yu Qiao, Hongyang Li

Modern autonomous driving system is characterized as modular tasks in sequential order, i.e., perception, prediction and planning. As sensors and hardware get improved, there is trending popularity to devise a system that can perform a wide diversity of tasks to fulfill higher-level intelligence. Contemporary approaches resort to either deploying standalone models for individual tasks, or designing a multi-task paradigm with separate heads. These might suffer from accumulative error or negative transfer effect. Instead, we argue that a favorable algorithm framework should be devised and optimized in pursuit of the ultimate goal, i.e. planning of the self-driving-car. Oriented at this goal, we revisit the key components within perception and prediction. We analyze each module and prioritize the tasks hierarchically, such that all these tasks contribute to planning (the goal). To this end, we introduce Unified Autonomous Driving (UniAD), the first comprehensive framework up-to-date that incorporates full-stack driving tasks in one network. It is exquisitely devised to leverage advantages of each module, and provide complementary feature abstractions for agent interaction from a global perspective. Tasks are communicated with unified query design to facilitate each other toward planning. We instantiate UniAD on the challenging nuScenes benchmark. With extensive ablations, the effectiveness of using such a philosophy is proven to surpass previous state-of-the-arts by a large margin in all aspects. The full suite of codebase and models would be available to facilitate future research in the community.

comment: Project page: https://opendrivelab.github.io/UniAD/

☆ Tracking by Associating Clips ECCV 2022

Sanghyun Woo, Kwanyong Park, Seoung Wug Oh, In So Kweon, Joon-Young Lee

The tracking-by-detection paradigm today has become the dominant method for multi-object tracking and works by detecting objects in each frame and then performing data association across frames. However, its sequential frame-wise matching property fundamentally suffers from the intermediate interruptions in a video, such as object occlusions, fast camera movements, and abrupt light changes. Moreover, it typically overlooks temporal information beyond the two frames for matching. In this paper, we investigate an alternative by treating object association as clip-wise matching. Our new perspective views a single long video sequence as multiple short clips, and then the tracking is performed both within and between the clips. The benefits of this new approach are two folds. First, our method is robust to tracking error accumulation or propagation, as the video chunking allows bypassing the interrupted frames, and the short clip tracking avoids the conventional error-prone long-term track memory management. Second, the multiple frame information is aggregated during the clip-wise matching, resulting in a more accurate long-range track association than the current frame-wise matching. Given the state-of-the-art tracking-by-detection tracker, QDTrack, we showcase how the tracking performance improves with our new tracking formulation. We evaluate our proposals on two tracking benchmarks, TAO and MOT17 that have complementary characteristics and challenges each other.

comment: ECCV 2022

☆ Bridging Images and Videos: A Simple Learning Framework for Large Vocabulary Video Object Detection ECCV 2022

Sanghyun Woo, Kwanyong Park, Seoung Wug Oh, In So Kweon, Joon-Young Lee

Scaling object taxonomies is one of the important steps toward a robust real-world deployment of recognition systems. We have faced remarkable progress in images since the introduction of the LVIS benchmark. To continue this success in videos, a new video benchmark, TAO, was recently presented. Given the recent encouraging results from both detection and tracking communities, we are interested in marrying those two advances and building a strong large vocabulary video tracker. However, supervisions in LVIS and TAO are inherently sparse or even missing, posing two new challenges for training the large vocabulary trackers. First, no tracking supervisions are in LVIS, which leads to inconsistent learning of detection (with LVIS and TAO) and tracking (only with TAO). Second, the detection supervisions in TAO are partial, which results in catastrophic forgetting of absent LVIS categories during video fine-tuning. To resolve these challenges, we present a simple but effective learning framework that takes full advantage of all available training data to learn detection and tracking while not losing any LVIS categories to recognize. With this new learning scheme, we show that consistent improvements of various large vocabulary trackers are capable, setting strong baseline results on the challenging TAO benchmarks.

comment: ECCV 2022

☆ Tackling Ambiguity with Images: Improved Multimodal Machine Translation and Contrastive Evaluation

Matthieu Futeral, Cordelia Schmid, Ivan Laptev, Benoît Sagot, Rachel Bawden

One of the major challenges of machine translation (MT) is ambiguity, which can in some cases be resolved by accompanying context such as an image. However, recent work in multimodal MT (MMT) has shown that obtaining improvements from images is challenging, limited not only by the difficulty of building effective cross-modal representations but also by the lack of specific evaluation and training data. We present a new MMT approach based on a strong text-only MT model, which uses neural adapters and a novel guided self-attention mechanism and which is jointly trained on both visual masking and MMT. We also release CoMMuTE, a Contrastive Multilingual Multimodal Translation Evaluation dataset, composed of ambiguous sentences and their possible translations, accompanied by disambiguating images corresponding to each translation. Our approach obtains competitive results over strong text-only models on standard English-to-French benchmarks and outperforms these baselines and state-of-the-art MMT systems with a large margin on our contrastive test set.

☆ Content Adaptive Latents and Decoder for Neural Image Compression ECCV 2022

Guanbo Pan, Guo Lu, Zhihao Hu, Dong Xu

In recent years, neural image compression (NIC) algorithms have shown powerful coding performance. However, most of them are not adaptive to the image content. Although several content adaptive methods have been proposed by updating the encoder-side components, the adaptability of both latents and the decoder is not well exploited. In this work, we propose a new NIC framework that improves the content adaptability on both latents and the decoder. Specifically, to remove redundancy in the latents, our content adaptive channel dropping (CACD) method automatically selects the optimal quality levels for the latents spatially and drops the redundant channels. Additionally, we propose the content adaptive feature transformation (CAFT) method to improve decoder-side content adaptability by extracting the characteristic information of the image content, which is then used to transform the features in the decoder side. Experimental results demonstrate that our proposed methods with the encoder-side updating algorithm achieve the state-of-the-art performance.

comment: V1 is accepted to ECCV 2022

☆ Image Segmentation-based Unsupervised Multiple Objects Discovery WACV 2023

Sandra Kara, Hejer Ammar, Florian Chabot, Quoc-Cuong Pham

Unsupervised object discovery aims to localize objects in images, while removing the dependence on annotations required by most deep learning-based methods. To address this problem, we propose a fully unsupervised, bottom-up approach, for multiple objects discovery. The proposed approach is a two-stage framework. First, instances of object parts are segmented by using the intra-image similarity between self-supervised local features. The second step merges and filters the object parts to form complete object instances. The latter is performed by two CNN models that capture semantic information on objects from the entire dataset. We demonstrate that the pseudo-labels generated by our method provide a better precision-recall trade-off than existing single and multiple objects discovery methods. In particular, we provide state-of-the-art results for both unsupervised class-agnostic object detection and unsupervised image segmentation.

comment: WACV 2023

☆ Efficient aggregation of face embeddings for decentralized face recognition deployments (extended version)

Philipp Hofer, Michael Roland, Philipp Schwarz, Renè Mayrhofer

Biometrics are one of the most privacy-sensitive data. Ubiquitous authentication systems with a focus on privacy favor decentralized approaches as they reduce potential attack vectors, both on a technical and organizational level. The gold standard is to let the user be in control of where their own data is stored, which consequently leads to a high variety of devices used. Moreover, in comparison with a centralized system, designs with higher end-user freedom often incur additional network overhead. Therefore, when using face recognition for biometric authentication, an efficient way to compare faces is important in practical deployments, because it reduces both network and hardware requirements that are essential to encourage device diversity. This paper proposes an efficient way to aggregate embeddings used for face recognition based on an extensive analysis on different datasets and the use of different aggregation strategies. As part of this analysis, a new dataset has been collected, which is available for research purposes. Our proposed method supports the construction of massively scalable, decentralized face recognition systems with a focus on both privacy and long-term usability.

☆ On the Applicability of Synthetic Data for Re-Identification AAAI23

Jérôme Rutinowski, Bhargav Vankayalapati, Nils Schwenzfeier, Maribel Acosta, Christopher Reining

This contribution demonstrates the feasibility of applying Generative Adversarial Networks (GANs) on images of EPAL pallet blocks for dataset enhancement in the context of re-identification. For many industrial applications of re-identification methods, datasets of sufficient volume would otherwise be unattainable in non-laboratory settings. Using a state-of-the-art GAN architecture, namely CycleGAN, images of pallet blocks rotated to their left-hand side were generated from images of visually centered pallet blocks, based on images of rotated pallet blocks that were recorded as part of a previously recorded and published dataset. In this process, the unique chipwood pattern of the pallet block surface structure was retained, only changing the orientation of the pallet block itself. By doing so, synthetic data for re-identification testing and training purposes was generated, in a manner that is distinct from ordinary data augmentation. In total, 1,004 new images of pallet blocks were generated. The quality of the generated images was gauged using a perspective classifier that was trained on the original images and then applied to the synthetic ones, comparing the accuracy between the two sets of images. The classification accuracy was 98% for the original images and 92% for the synthetic images. In addition, the generated images were also used in a re-identification task, in order to re-identify original images based on synthetic ones. The accuracy in this scenario was up to 88% for synthetic images, compared to 96% for original images. Through this evaluation, it is established, whether or not a generated pallet block image closely resembles its original counterpart.

comment: Accepted as a non-archival paper in AAAI23 workshop AI2SE

☆ Visual Transformers for Primates Classification and Covid Detection

Steffen Illium, Robert Müller, Andreas Sedlmeier, Claudia-Linnhoff Popien

We apply the vision transformer, a deep machine learning model build around the attention mechanism, on mel-spectrogram representations of raw audio recordings. When adding mel-based data augmentation techniques and sample-weighting, we achieve comparable performance on both (PRS and CCS challenge) tasks of ComParE21, outperforming most single model baselines. We further introduce overlapping vertical patching and evaluate the influence of parameter configurations. Index Terms: audio classification, attention, mel-spectrogram, unbalanced data-sets, computational paralinguistics

☆ Computer Vision Methods for Automating Turbot Fish Cutting

Fernando Martin-Rodriguez, Fernando Isasi-de-Vicente, Monica Fernandez-Barciela

This paper is about the design of an automated machine to cut turbot fish specimens. Machine vision is a key part of this project as it is used to compute a cutting curve for the specimen head. This task is impossible to be carried out by mechanical means. Machine vision is used to detect head boundary and a robot is used to cut the head. Binarization and mathematical morphology are used to detect fish boundary and this boundary is subsequently analyzed (using Hough transform and convex hull) to detect key points and thus defining the cutting curve. Afterwards, mechanical systems are used to slice fish to get an easy presentation for end consumer (as fish fillets than can be easily marketed and consumed).

comment: 5 pages, 11 figurs. Derived from conference publication: https://upcommons.upc.edu/handle/2117/77428

☆ End to End Generative Meta Curriculum Learning For Medical Data Augmentation

Meng Li, Brian Lovell

Current medical image synthetic augmentation techniques rely on intensive use of generative adversarial networks (GANs). However, the nature of GAN architecture leads to heavy computational resources to produce synthetic images and the augmentation process requires multiple stages to complete. To address these challenges, we introduce a novel generative meta curriculum learning method that trains the task-specific model (student) end-to-end with only one additional teacher model. The teacher learns to generate curriculum to feed into the student model for data augmentation and guides the student to improve performance in a meta-learning style. In contrast to the generator and discriminator in GAN, which compete with each other, the teacher and student collaborate to improve the student's performance on the target tasks. Extensive experiments on the histopathology datasets show that leveraging our framework results in significant and consistent improvements in classification performance.

☆ An Information-Theoretic Approach to Transferability in Task Transfer Learning

Yajie Bao, Yang Li, Shao-Lun Huang, Lin Zhang, Lizhong Zheng, Amir Zamir, Leonidas Guibas

Task transfer learning is a popular technique in image processing applications that uses pre-trained models to reduce the supervision cost of related tasks. An important question is to determine task transferability, i.e. given a common input domain, estimating to what extent representations learned from a source task can help in learning a target task. Typically, transferability is either measured experimentally or inferred through task relatedness, which is often defined without a clear operational meaning. In this paper, we present a novel metric, H-score, an easily-computable evaluation function that estimates the performance of transferred representations from one task to another in classification problems using statistical and information theoretic principles. Experiments on real image data show that our metric is not only consistent with the empirical transferability measurement, but also useful to practitioners in applications such as source model selection and task transfer curriculum learning.

☆ RepMode: Learning to Re-parameterize Diverse Experts for Subcellular Structure Prediction

Donghao Zhou, Chunbin Gu, Junde Xu, Furui Liu, Qiong Wang, Guangyong Chen, Pheng-Ann Heng

In subcellular biological research, fluorescence staining is a key technique to reveal the locations and morphology of subcellular structures. However, fluorescence staining is slow, expensive, and harmful to cells. In this paper, we treat it as a deep learning task termed subcellular structure prediction (SSP), aiming to predict the 3D fluorescent images of multiple subcellular structures from a 3D transmitted-light image. Unfortunately, due to the limitations of current biotechnology, each image is partially labeled in SSP. Besides, naturally, the subcellular structures vary considerably in size, which causes the multi-scale issue in SSP. However, traditional solutions can not address SSP well since they organize network parameters inefficiently and inflexibly. To overcome these challenges, we propose Re-parameterizing Mixture-of-Diverse-Experts (RepMode), a network that dynamically organizes its parameters with task-aware priors to handle specified single-label prediction tasks of SSP. In RepMode, the Mixture-of-Diverse-Experts (MoDE) block is designed to learn the generalized parameters for all tasks, and gating re-parameterization (GatRep) is performed to generate the specialized parameters for each task, by which RepMode can maintain a compact practical topology exactly like a plain network, and meanwhile achieves a powerful theoretical topology. Comprehensive experiments show that RepMode outperforms existing methods on ten of twelve prediction tasks of SSP and achieves state-of-the-art overall performance.

comment: 16 pages, 10 figures

☆ VoronoiPatches: Evaluating A New Data Augmentation Method

Steffen Illium, Gretchen Griffin, Michael Kölle, Maximilian Zorn, Jonas Nüßlein, Claudia Linnhoff-Popien

Overfitting is a problem in Convolutional Neural Networks (CNN) that causes poor generalization of models on unseen data. To remediate this problem, many new and diverse data augmentation methods (DA) have been proposed to supplement or generate more training data, and thereby increase its quality. In this work, we propose a new data augmentation algorithm: VoronoiPatches (VP). We primarily utilize non-linear recombination of information within an image, fragmenting and occluding small information patches. Unlike other DA methods, VP uses small convex polygon-shaped patches in a random layout to transport information around within an image. Sudden transitions created between patches and the original image can, optionally, be smoothed. In our experiments, VP outperformed current DA methods regarding model variance and overfitting tendencies. We demonstrate data augmentation utilizing non-linear re-combination of information within images, and non-orthogonal shapes and structures improves CNN model robustness on unseen data.

☆ OBMO: One Bounding Box Multiple Objects for Monocular 3D Object Detection

Chenxi Huang, Tong He, Haidong Ren, Wenxiao Wang, Binbin Lin, Deng Cai

Compared to typical multi-sensor systems, monocular 3D object detection has attracted much attention due to its simple configuration. However, there is still a significant gap between LiDAR-based and monocular-based methods. In this paper, we find that the ill-posed nature of monocular imagery can lead to depth ambiguity. Specifically, objects with different depths can appear with the same bounding boxes and similar visual features in the 2D image. Unfortunately, the network cannot accurately distinguish different depths from such non-discriminative visual features, resulting in unstable depth training. To facilitate depth learning, we propose a simple yet effective plug-and-play module, One Bounding Box Multiple Objects (OBMO). Concretely, we add a set of suitable pseudo labels by shifting the 3D bounding box along the viewing frustum. To constrain the pseudo-3D labels to be reasonable, we carefully design two label scoring strategies to represent their quality. In contrast to the original hard depth labels, such soft pseudo labels with quality scores allow the network to learn a reasonable depth range, boosting training stability and thus improving final performance. Extensive experiments on KITTI and Waymo benchmarks show that our method significantly improves state-of-the-art monocular 3D detectors by a significant margin (The improvements under the moderate setting on KITTI validation set are $\mathbf{1.82\sim 10.91\%}$ mAP in BEV and $\mathbf{1.18\sim 9.36\%}$ mAP in 3D}. Codes have been released at https://github.com/mrsempress/OBMO.

comment: 9 pages, 9 figures

☆ Benchmarking Spatial Relationships in Text-to-Image Generation

Tejas Gokhale, Hamid Palangi, Besmira Nushi, Vibhav Vineet, Eric Horvitz, Ece Kamar, Chitta Baral, Yezhou Yang

Spatial understanding is a fundamental aspect of computer vision and integral for human-level reasoning about images, making it an important component for grounded language understanding. While recent large-scale text-to-image synthesis (T2I) models have shown unprecedented improvements in photorealism, it is unclear whether they have reliable spatial understanding capabilities. We investigate the ability of T2I models to generate correct spatial relationships among objects and present VISOR, an evaluation metric that captures how accurately the spatial relationship described in text is generated in the image. To benchmark existing models, we introduce a large-scale challenge dataset SR2D that contains sentences describing two objects and the spatial relationship between them. We construct and harness an automated evaluation pipeline that employs computer vision to recognize objects and their spatial relationships, and we employ it in a large-scale evaluation of T2I models. Our experiments reveal a surprising finding that, although recent state-of-the-art T2I models exhibit high image quality, they are severely limited in their ability to generate multiple objects or the specified spatial relations such as left/right/above/below. Our analyses demonstrate several biases and artifacts of T2I models such as the difficulty with generating multiple objects, a bias towards generating the first object mentioned, spatially inconsistent outputs for equivalent relationships, and a correlation between object co-occurrence and spatial understanding capabilities. We conduct a human study that shows the alignment between VISOR and human judgment about spatial understanding. We offer the SR2D dataset and the VISOR metric to the community in support of T2I spatial reasoning research.

comment: preprint

☆ Calibrating Deep Neural Networks using Explicit Regularisation and Dynamic Data Pruning WACV

Ramya Hebbalaguppe, Rishabh Patra, Tirtharaj Dash, Gautam Shroff, Lovekesh Vig

Deep neural networks (DNN) are prone to miscalibrated predictions, often exhibiting a mismatch between the predicted output and the associated confidence scores. Contemporary model calibration techniques mitigate the problem of overconfident predictions by pushing down the confidence of the winning class while increasing the confidence of the remaining classes across all test samples. However, from a deployment perspective, an ideal model is desired to (i) generate well-calibrated predictions for high-confidence samples with predicted probability say >0.95, and (ii) generate a higher proportion of legitimate high-confidence samples. To this end, we propose a novel regularization technique that can be used with classification losses, leading to state-of-the-art calibrated predictions at test time; From a deployment standpoint in safety-critical applications, only high-confidence samples from a well-calibrated model are of interest, as the remaining samples have to undergo manual inspection. Predictive confidence reduction of these potentially ``high-confidence samples'' is a downside of existing calibration approaches. We mitigate this by proposing a dynamic train-time data pruning strategy that prunes low-confidence samples every few epochs, providing an increase in "confident yet calibrated samples". We demonstrate state-of-the-art calibration performance across image classification benchmarks, reducing training time without much compromise in accuracy. We provide insights into why our dynamic pruning strategy that prunes low-confidence training samples leads to an increase in high-confidence samples at test time.

comment: The paper is accepted at Winter Conference on applications of Computer Vision (IEEE WACV) in algorithms tracks. 8 pages Main paper; 3 pages supplementary material

☆ Are Deep Neural Networks SMARTer than Second Graders?

Anoop Cherian, Kuan-Chuan Peng, Suhas Lohit, Kevin Smith, Joshua B. Tenenbaum

Recent times have witnessed an increasing number of applications of deep neural networks towards solving tasks that require superior cognitive abilities, e.g., playing Go, generating art, question answering (such as ChatGPT), etc. Such a dramatic progress raises the question: how generalizable are neural networks in solving problems that demand broad skills? To answer this question, we propose SMART: a Simple Multimodal Algorithmic Reasoning Task and the associated SMART-101 dataset, for evaluating the abstraction, deduction, and generalization abilities of neural networks in solving visuo-linguistic puzzles designed specifically for children in the 6-8 age group. Our dataset consists of 101 unique puzzles; each puzzle comprises a picture and a question, and their solution needs a mix of several elementary skills, including arithmetic, algebra, and spatial reasoning, among others. To scale our dataset towards training deep neural networks, we programmatically generate entirely new instances for each puzzle while retaining their solution algorithm. To benchmark the performance on the SMART-101 dataset, we propose a vision and language meta-learning model using varied state-of-the-art backbone neural networks. Our experiments reveal that while powerful deep models offer reasonable performances on puzzles that they are trained on, they are not better than random accuracy when analyzed for generalization. We also evaluate the recent ChatGPT large language model on a subset of our dataset and find that while ChatGPT produces convincing reasoning abilities, the answers are often incorrect.

☆ Multi-Reference Image Super-Resolution: A Posterior Fusion Approach

Ke Zhao, Haining Tan, Tsz Fung Yau

Reference-based Super-resolution (RefSR) approaches have recently been proposed to overcome the ill-posed problem of image super-resolution by providing additional information from a high-resolution image. Multi-reference super-resolution extends this approach by allowing more information to be incorporated. This paper proposes a 2-step-weighting posterior fusion approach to combine the outputs of RefSR models with multiple references. Extensive experiments on the CUFED5 dataset demonstrate that the proposed methods can be applied to various state-of-the-art RefSR models to get a consistent improvement in image quality.

☆ Texture Representation via Analysis and Synthesis with Generative Adversarial Networks

Jue Lin, Gaurav Sharma, Thrasyvoulos N. Pappas

We investigate data-driven texture modeling via analysis and synthesis with generative adversarial networks. For network training and testing, we have compiled a diverse set of spatially homogeneous textures, ranging from stochastic to regular. We adopt StyleGAN3 for synthesis and demonstrate that it produces diverse textures beyond those represented in the training data. For texture analysis, we propose GAN inversion using a novel latent domain reconstruction consistency criterion for synthesized textures, and iterative refinement with Gramian loss for real textures. We propose perceptual procedures for evaluating network capabilities, exploring the global and local behavior of latent space trajectories, and comparing with existing texture analysis-synthesis techniques.

☆ Benchmarking person re-identification datasets and approaches for practical real-world implementations

Jose Huaman, Felix O. Sumari, Luigy Machaca, Esteban Clua, Joris Guerin

Recently, Person Re-Identification (Re-ID) has received a lot of attention. Large datasets containing labeled images of various individuals have been released, allowing researchers to develop and test many successful approaches. However, when such Re-ID models are deployed in new cities or environments, the task of searching for people within a network of security cameras is likely to face an important domain shift, thus resulting in decreased performance. Indeed, while most public datasets were collected in a limited geographic area, images from a new city present different features (e.g., people's ethnicity and clothing style, weather, architecture, etc.). In addition, the whole frames of the video streams must be converted into cropped images of people using pedestrian detection models, which behave differently from the human annotators who created the dataset used for training. To better understand the extent of this issue, this paper introduces a complete methodology to evaluate Re-ID approaches and training datasets with respect to their suitability for unsupervised deployment for live operations. This method is used to benchmark four Re-ID approaches on three datasets, providing insight and guidelines that can help to design better Re-ID pipelines in the future.

comment: This paper is the extended version of our short paper accepted in VISAPP - 2023

☆ Flareon: Stealthy any2any Backdoor Injection via Poisoned Augmentation

Tianrui Qin, Xianghuan He, Xitong Gao, Yiren Zhao, Kejiang Ye, Cheng-Zhong Xu

Open software supply chain attacks, once successful, can exact heavy costs in mission-critical applications. As open-source ecosystems for deep learning flourish and become increasingly universal, they present attackers previously unexplored avenues to code-inject malicious backdoors in deep neural network models. This paper proposes Flareon, a small, stealthy, seemingly harmless code modification that specifically targets the data augmentation pipeline with motion-based triggers. Flareon neither alters ground-truth labels, nor modifies the training loss objective, nor does it assume prior knowledge of the victim model architecture, training data, and training hyperparameters. Yet, it has a surprisingly large ramification on training -- models trained under Flareon learn powerful target-conditional (or "any2any") backdoors. The resulting models can exhibit high attack success rates for any target choices and better clean accuracies than backdoor attacks that not only seize greater control, but also assume more restrictive attack capabilities. We also demonstrate the effectiveness of Flareon against recent defenses. Flareon is fully open-source and available online to the deep learning community: https://github.com/lafeat/flareon.

☆ Conditioned Generative Transformers for Histopathology Image Synthetic Augmentation

Meng Li, Chaoyi Li, Can Peng, Brian Lovell

Deep learning networks have demonstrated state-of-the-art performance on medical image analysis tasks. However, the majority of the works rely heavily on abundantly labeled data, which necessitates extensive involvement of domain experts. Vision transformer (ViT) based generative adversarial networks (GANs) recently demonstrated superior potential in general image synthesis, yet are less explored for histopathology images. In this paper, we address these challenges by proposing a pure ViT-based conditional GAN model for histopathology image synthetic augmentation. To alleviate training instability and improve generation robustness, we first introduce a conditioned class projection method to facilitate class separation. We then implement a multi-loss weighing function to dynamically balance the losses between classification tasks. We further propose a selective augmentation mechanism to actively choose the appropriate generated images and bring additional performance improvements. Extensive experiments on the histopathology datasets show that leveraging our synthetic augmentation framework results in significant and consistent improvements in classification performance.

☆ Domain Generalization with Correlated Style Uncertainty

Zheyuan Zhang, Bin Wang, Debesh Jha, Ugur Demir, Ulas Bagci

Though impressive success has been witnessed in computer vision, deep learning still suffers from the domain shift challenge when the target domain for testing and the source domain for training do not share an identical distribution. To address this, domain generalization approaches intend to extract domain invariant features that can lead to a more robust model. Hence, increasing the source domain diversity is a key component of domain generalization. Style augmentation takes advantage of instance-specific feature statistics containing informative style characteristics to synthetic novel domains. However, all previous works ignored the correlation between different feature channels or only limited the style augmentation through linear interpolation. In this work, we propose a novel augmentation method, called \textit{Correlated Style Uncertainty (CSU)}, to go beyond the linear interpolation of style statistic space while preserving the essential correlation information. We validate our method's effectiveness by extensive experiments on multiple cross-domain classification tasks, including widely used PACS, Office-Home, Camelyon17 datasets and the Duke-Market1501 instance retrieval task and obtained significant margin improvements over the state-of-the-art methods. The source code is available for public use.

comment: Code is available after peer review

☆ MM-3DScene: 3D Scene Understanding by Customizing Masked Modeling with Informative-Preserved Reconstruction and Self-Distilled Consistency

Mingye Xu, Mutian Xu, Tong He, Wanli Ouyang, Yali Wang, Xiaoguang Han, Yu Qiao

Masked Modeling (MM) has demonstrated widespread success in various vision challenges, by reconstructing masked visual patches. Yet, applying MM for large-scale 3D scenes remains an open problem due to the data sparsity and scene complexity. The conventional random masking paradigm used in 2D images often causes a high risk of ambiguity when recovering the masked region of 3D scenes. To this end, we propose a novel informative-preserved reconstruction, which explores local statistics to discover and preserve the representative structured points, effectively enhancing the pretext masking task for 3D scene understanding. Integrated with a progressive reconstruction manner, our method can concentrate on modeling regional geometry and enjoy less ambiguity for masked reconstruction. Besides, such scenes with progressive masking ratios can also serve to self-distill their intrinsic spatial consistency, requiring to learn the consistent representations from unmasked areas. By elegantly combining informative-preserved reconstruction on masked areas and consistency self-distillation from unmasked areas, a unified framework called MM-3DScene is yielded. We conduct comprehensive experiments on a host of downstream tasks. The consistent improvement (e.g., +6.1 mAP@0.5 on object detection and +2.2% mIoU on semantic segmentation) demonstrates the superiority of our approach.

☆ Robust and Resource-efficient Machine Learning Aided Viewport Prediction in Virtual Reality

Yuang Jiang, Konstantinos Poularakis, Diego Kiedanski, Sastry Kompella, Leandros Tassiulas

360-degree panoramic videos have gained considerable attention in recent years due to the rapid development of head-mounted displays (HMDs) and panoramic cameras. One major problem in streaming panoramic videos is that panoramic videos are much larger in size compared to traditional ones. Moreover, the user devices are often in a wireless environment, with limited battery, computation power, and bandwidth. To reduce resource consumption, researchers have proposed ways to predict the users' viewports so that only part of the entire video needs to be transmitted from the server. However, the robustness of such prediction approaches has been overlooked in the literature: it is usually assumed that only a few models, pre-trained on past users' experiences, are applied for prediction to all users. We observe that those pre-trained models can perform poorly for some users because they might have drastically different behaviors from the majority, and the pre-trained models cannot capture the features in unseen videos. In this work, we propose a novel meta learning based viewport prediction paradigm to alleviate the worst prediction performance and ensure the robustness of viewport prediction. This paradigm uses two machine learning models, where the first model predicts the viewing direction, and the second model predicts the minimum video prefetch size that can include the actual viewport. We first train two meta models so that they are sensitive to new training data, and then quickly adapt them to users while they are watching the videos. Evaluation results reveal that the meta models can adapt quickly to each user, and can significantly increase the prediction accuracy, especially for the worst-performing predictions.

comment: Accepted for publication in 2022 IEEE International Conference on Big Data (IEEE BigData 2022)

☆ METEOR Guided Divergence for Video Captioning

Daniel Lukas Rothenpieler, Shahin Amiriparian

Automatic video captioning aims for a holistic visual scene understanding. It requires a mechanism for capturing temporal context in video frames and the ability to comprehend the actions and associations of objects in a given timeframe. Such a system should additionally learn to abstract video sequences into sensible representations as well as to generate natural written language. While the majority of captioning models focus solely on the visual inputs, little attention has been paid to the audiovisual modality. To tackle this issue, we propose a novel two-fold approach. First, we implement a reward-guided KL Divergence to train a video captioning model which is resilient towards token permutations. Second, we utilise a Bi-Modal Hierarchical Reinforcement Learning (BMHRL) Transformer architecture to capture long-term temporal dependencies of the input data as a foundation for our hierarchical captioning module. Using our BMHRL, we show the suitability of the HRL agent in the generation of content-complete and grammatically sound sentences by achieving $4.91$, $2.23$, and $10.80$ in BLEU3, BLEU4, and METEOR scores, respectively on the ActivityNet Captions dataset. Finally, we make our BMHRL framework and trained models publicly available for users and developers at https://github.com/d-rothen/bmhrl.

☆ Local Differential Privacy Image Generation Using Flow-based Deep Generative Models

Hisaichi Shibata, Shouhei Hanaoka, Yang Cao, Masatoshi Yoshikawa, Tomomi Takenaga, Yukihiro Nomura, Naoto Hayashi, Osamu Abe

Diagnostic radiologists need artificial intelligence (AI) for medical imaging, but access to medical images required for training in AI has become increasingly restrictive. To release and use medical images, we need an algorithm that can simultaneously protect privacy and preserve pathologies in medical images. To develop such an algorithm, here, we propose DP-GLOW, a hybrid of a local differential privacy (LDP) algorithm and one of the flow-based deep generative models (GLOW). By applying a GLOW model, we disentangle the pixelwise correlation of images, which makes it difficult to protect privacy with straightforward LDP algorithms for images. Specifically, we map images onto the latent vector of the GLOW model, each element of which follows an independent normal distribution, and we apply the Laplace mechanism to the latent vector. Moreover, we applied DP-GLOW to chest X-ray images to generate LDP images while preserving pathologies.

☆ Privacy-Protecting Behaviours of Risk Detection in People with Dementia using Videos

Pratik K. Mishra, Andrea Iaboni, Bing Ye, Kristine Newman, Alex Mihailidis, Shehroz S. Khan

People living with dementia often exhibit behavioural and psychological symptoms of dementia that can put their and others' safety at risk. Existing video surveillance systems in long-term care facilities can be used to monitor such behaviours of risk to alert the staff to prevent potential injuries or death in some cases. However, these behaviours of risk events are heterogeneous and infrequent in comparison to normal events. Moreover, analyzing raw videos can also raise privacy concerns. In this paper, we present two novel privacy-protecting video-based anomaly detection approaches to detect behaviours of risks in people with dementia. We either extracted body pose information as skeletons and use semantic segmentation masks to replace multiple humans in the scene with their semantic boundaries. Our work differs from most existing approaches for video anomaly detection that focus on appearance-based features, which can put the privacy of a person at risk and is also susceptible to pixel-based noise, including illumination and viewing direction. We used anonymized videos of normal activities to train customized spatio-temporal convolutional autoencoders and identify behaviours of risk as anomalies. We show our results on a real-world study conducted in a dementia care unit with patients with dementia, containing approximately 21 hours of normal activities data for training and 9 hours of data containing normal and behaviours of risk events for testing. We compared our approaches with the original RGB videos and obtained an equivalent area under the receiver operating characteristic curve performance of 0.807 for the skeleton-based approach and 0.823 for the segmentation mask-based approach. This is one of the first studies to incorporate privacy for the detection of behaviours of risks in people with dementia.

☆ High-Throughput, High-Performance Deep Learning-Driven Light Guide Plate Surface Visual Quality Inspection Tailored for Real-World Manufacturing Environments

Carol Xu, Mahmoud Famouri, Gautam Bathla, Mohammad Javad Shafiee, Alexander Wong

Light guide plates are essential optical components widely used in a diverse range of applications ranging from medical lighting fixtures to back-lit TV displays. In this work, we introduce a fully-integrated, high-throughput, high-performance deep learning-driven workflow for light guide plate surface visual quality inspection (VQI) tailored for real-world manufacturing environments. To enable automated VQI on the edge computing within the fully-integrated VQI system, a highly compact deep anti-aliased attention condenser neural network (which we name LightDefectNet) tailored specifically for light guide plate surface defect detection in resource-constrained scenarios was created via machine-driven design exploration with computational and "best-practices" constraints as well as L_1 paired classification discrepancy loss. Experiments show that LightDetectNet achieves a detection accuracy of ~98.2% on the LGPSDD benchmark while having just 770K parameters (~33X and ~6.9X lower than ResNet-50 and EfficientNet-B0, respectively) and ~93M FLOPs (~88X and ~8.4X lower than ResNet-50 and EfficientNet-B0, respectively) and ~8.8X faster inference speed than EfficientNet-B0 on an embedded ARM processor. As such, the proposed deep learning-driven workflow, integrated with the aforementioned LightDefectNet neural network, is highly suited for high-throughput, high-performance light plate surface VQI within real-world manufacturing environments.

comment: arXiv admin note: substantial text overlap with arXiv:2204.11765

☆ CHAIRS: Towards Full-Body Articulated Human-Object Interaction

Nan Jiang, Tengyu Liu, Zhexuan Cao, Jieming Cui, Yixin Chen, He Wang, Yixin Zhu, Siyuan Huang

Fine-grained capturing of 3D HOI boosts human activity understanding and facilitates downstream visual tasks, including action recognition, holistic scene reconstruction, and human motion synthesis. Despite its significance, existing works mostly assume that humans interact with rigid objects using only a few body parts, limiting their scope. In this paper, we address the challenging problem of f-AHOI, wherein the whole human bodies interact with articulated objects, whose parts are connected by movable joints. We present CHAIRS, a large-scale motion-captured f-AHOI dataset, consisting of 16.2 hours of versatile interactions between 46 participants and 81 articulated and rigid sittable objects. CHAIRS provides 3D meshes of both humans and articulated objects during the entire interactive process, as well as realistic and physically plausible full-body interactions. We show the value of CHAIRS with object pose estimation. By learning the geometrical relationships in HOI, we devise the very first model that leverage human pose estimation to tackle the estimation of articulated object poses and shapes during whole-body interactions. Given an image and an estimated human pose, our model first reconstructs the pose and shape of the object, then optimizes the reconstruction according to a learned interaction prior. Under both evaluation settings (e.g., with or without the knowledge of objects' geometries/structures), our model significantly outperforms baselines. We hope CHAIRS will promote the community towards finer-grained interaction understanding. We will make the data/code publicly available.

☆ Temporal Output Discrepancy for Loss Estimation-based Active Learning ICCV 2021

Siyu Huang, Tianyang Wang, Haoyi Xiong, Bihan Wen, Jun Huan, Dejing Dou

While deep learning succeeds in a wide range of tasks, it highly depends on the massive collection of annotated data which is expensive and time-consuming. To lower the cost of data annotation, active learning has been proposed to interactively query an oracle to annotate a small proportion of informative samples in an unlabeled dataset. Inspired by the fact that the samples with higher loss are usually more informative to the model than the samples with lower loss, in this paper we present a novel deep active learning approach that queries the oracle for data annotation when the unlabeled sample is believed to incorporate high loss. The core of our approach is a measurement Temporal Output Discrepancy (TOD) that estimates the sample loss by evaluating the discrepancy of outputs given by models at different optimization steps. Our theoretical investigation shows that TOD lower-bounds the accumulated sample loss thus it can be used to select informative unlabeled samples. On basis of TOD, we further develop an effective unlabeled data sampling strategy as well as an unsupervised learning criterion for active learning. Due to the simplicity of TOD, our methods are efficient, flexible, and task-agnostic. Extensive experimental results demonstrate that our approach achieves superior performances than the state-of-the-art active learning methods on image classification and semantic segmentation tasks. In addition, we show that TOD can be utilized to select the best model of potentially the highest testing accuracy from a pool of candidate models.

comment: Accepted for IEEE Transactions on Neural Networks and Learning Systems, 2022. Journal extension of ICCV 2021 [arXiv:2107.14153]

☆ Open-Vocabulary Temporal Action Detection with Off-the-Shelf Image-Text Features

Vivek Rathod, Bryan Seybold, Sudheendra Vijayanarasimhan, Austin Myers, Xiuye Gu, Vighnesh Birodkar, David A. Ross

Detecting actions in untrimmed videos should not be limited to a small, closed set of classes. We present a simple, yet effective strategy for open-vocabulary temporal action detection utilizing pretrained image-text co-embeddings. Despite being trained on static images rather than videos, we show that image-text co-embeddings enable openvocabulary performance competitive with fully-supervised models. We show that the performance can be further improved by ensembling the image-text features with features encoding local motion, like optical flow based features, or other modalities, like audio. In addition, we propose a more reasonable open-vocabulary evaluation setting for the ActivityNet data set, where the category splits are based on similarity rather than random assignment.

☆ Video Segmentation Learning Using Cascade Residual Convolutional Neural Network

Daniel F. S. Santos, Rafael G. Pires, Danilo Colombo, João P. Papa

Video segmentation consists of a frame-by-frame selection process of meaningful areas related to foreground moving objects. Some applications include traffic monitoring, human tracking, action recognition, efficient video surveillance, and anomaly detection. In these applications, it is not rare to face challenges such as abrupt changes in weather conditions, illumination issues, shadows, subtle dynamic background motions, and also camouflage effects. In this work, we address such shortcomings by proposing a novel deep learning video segmentation approach that incorporates residual information into the foreground detection learning process. The main goal is to provide a method capable of generating an accurate foreground detection given a grayscale video. Experiments conducted on the Change Detection 2014 and on the private dataset PetrobrasROUTES from Petrobras support the effectiveness of the proposed approach concerning some state-of-the-art video segmentation techniques, with overall F-measures of $\mathbf{0.9535}$ and $\mathbf{0.9636}$ in the Change Detection 2014 and PetrobrasROUTES datasets, respectively. Such a result places the proposed technique amongst the top 3 state-of-the-art video segmentation methods, besides comprising approximately seven times less parameters than its top one counterpart.

comment: Published in: 2019 32nd SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI). arXiv admin note: text overlap with arXiv:2212.10417

☆ C2F-TCN: A Framework for Semi and Fully Supervised Temporal Action Segmentation

Dipika Singhania, Rahul Rahaman, Angela Yao

Temporal action segmentation tags action labels for every frame in an input untrimmed video containing multiple actions in a sequence. For the task of temporal action segmentation, we propose an encoder-decoder-style architecture named C2F-TCN featuring a "coarse-to-fine" ensemble of decoder outputs. The C2F-TCN framework is enhanced with a novel model agnostic temporal feature augmentation strategy formed by the computationally inexpensive strategy of the stochastic max-pooling of segments. It produces more accurate and well-calibrated supervised results on three benchmark action segmentation datasets. We show that the architecture is flexible for both supervised and representation learning. In line with this, we present a novel unsupervised way to learn frame-wise representation from C2F-TCN. Our unsupervised learning approach hinges on the clustering capabilities of the input features and the formation of multi-resolution features from the decoder's implicit structure. Further, we provide the first semi-supervised temporal action segmentation results by merging representation learning with conventional supervised learning. Our semi-supervised learning scheme, called ``Iterative-Contrastive-Classify (ICC)'', progressively improves in performance with more labeled data. The ICC semi-supervised learning in C2F-TCN, with 40% labeled videos, performs similar to fully supervised counterparts.

comment: arXiv admin note: text overlap with arXiv:2112.01402

☆ Empirical Analysis of Limits for Memory Distance in Recurrent Neural Networks

Steffen Illium, Thore Schillman, Robert Müller, Thomas Gabor, Claudia Linnhoff-Popien

Common to all different kinds of recurrent neural networks (RNNs) is the intention to model relations between data points through time. When there is no immediate relationship between subsequent data points (like when the data points are generated at random, e.g.), we show that RNNs are still able to remember a few data points back into the sequence by memorizing them by heart using standard backpropagation. However, we also show that for classical RNNs, LSTM and GRU networks the distance of data points between recurrent calls that can be reproduced this way is highly limited (compared to even a loose connection between data points) and subject to various constraints imposed by the type and size of the RNN in question. This implies the existence of a hard limit (way below the information-theoretic one) for the distance between related data points within which RNNs are still able to recognize said relation.

☆ Scene-aware Egocentric 3D Human Pose Estimation

Jian Wang, Lingjie Liu, Weipeng Xu, Kripasindhu Sarkar, Diogo Luvizon, Christian Theobalt

Egocentric 3D human pose estimation with a single head-mounted fisheye camera has recently attracted attention due to its numerous applications in virtual and augmented reality. Existing methods still struggle in challenging poses where the human body is highly occluded or is closely interacting with the scene. To address this issue, we propose a scene-aware egocentric pose estimation method that guides the prediction of the egocentric pose with scene constraints. To this end, we propose an egocentric depth estimation network to predict the scene depth map from a wide-view egocentric fisheye camera while mitigating the occlusion of the human body with a depth-inpainting network. Next, we propose a scene-aware pose estimation network that projects the 2D image features and estimated depth map of the scene into a voxel space and regresses the 3D pose with a V2V network. The voxel-based feature representation provides the direct geometric connection between 2D image features and scene geometry, and further facilitates the V2V network to constrain the predicted pose based on the estimated scene geometry. To enable the training of the aforementioned networks, we also generated a synthetic dataset, called EgoGTA, and an in-the-wild dataset based on EgoPW, called EgoPW-Scene. The experimental results of our new evaluation sequences show that the predicted 3D egocentric poses are accurate and physically plausible in terms of human-scene interaction, demonstrating that our method outperforms the state-of-the-art methods both quantitatively and qualitatively.

♻ ☆ Clover: Towards A Unified Video-Language Alignment and Fusion Model

Jingjia Huang, Yinan Li, Jiashi Feng, Xinglong Wu, Xiaoshuai Sun, Rongrong Ji

Building a universal Video-Language model for solving various video understanding tasks (\emph{e.g.}, text-video retrieval, video question answering) is an open challenge to the machine learning field. Towards this goal, most recent works build the model by stacking uni-modal and cross-modal feature encoders and train it with pair-wise contrastive pre-text tasks. Though offering attractive generality, the resulted models have to compromise between efficiency and performance. They mostly adopt different architectures to deal with different downstream tasks. We find this is because the pair-wise training cannot well \emph{align} and \emph{fuse} features from different modalities. We then introduce \textbf{Clover}\textemdash a Correlated Video-Language pre-training method\textemdash towards a universal Video-Language model for solving multiple video understanding tasks with neither performance nor efficiency compromise. It improves cross-modal feature alignment and fusion via a novel tri-modal alignment pre-training task. Additionally, we propose to enhance the tri-modal alignment via incorporating learning from semantic masked samples and a new pair-wise ranking loss. Clover establishes new state-of-the-arts on multiple downstream tasks, including three retrieval tasks for both zero-shot and fine-tuning settings, and eight video question answering tasks. Codes and pre-trained models will be released at \url{https://github.com/LeeYN-43/Clover}.

comment: Update Tri-modal Alignment task

♻ ☆ DALL-E for Detection: Language-driven Compositional Image Synthesis for Object Detection

Yunhao Ge, Jiashu Xu, Brian Nlong Zhao, Neel Joshi, Laurent Itti, Vibhav Vineet

We propose a new paradigm to automatically generate training data with accurate labels at scale using the text-toimage synthesis frameworks (e.g., DALL-E, Stable Diffusion, etc.). The proposed approach decouples training data generation into foreground object mask generation and background (context) image generation. For foreground object mask generation, we use a simple textual template with object class name as input to DALL-E to generate a diverse set of foreground images. A foreground-background segmentation algorithm is then used to generate foreground object masks. Next, in order to generate context images, first a language description of the context is generated by applying an image captioning method on a small set of images representing the context. These language descriptions are then used to generate diverse sets of context images using the DALL-E framework. These are then composited with object masks generated in the first step to provide an augmented training set for a classifier. We demonstrate the advantages of our approach on four object detection datasets including on Pascal VOC and COCO object detection tasks. Furthermore, we also highlight the compositional nature of our data generation approach on out-of-distribution and zero-shot data generation scenarios.

comment: v2 version, update structure (add foreground generation, stable diffusion), add more experiments

♻ ☆ GAUSS: Guided Encoder-Decoder Architecture for Hyperspectral Unmixing with Spatial Smoothness

Yasiru Ranasinghe, Kavinga Weerasooriya, Roshan Godaliyadda, Vijitha Herath, Parakrama Ekanayake, Dhananjaya Jayasundara, Lakshitha Ramanayake, Neranjan Senarath, Dulantha Wickramasinghe

In recent hyperspectral unmixing (HU) literature, the application of deep learning (DL) has become more prominent, especially with the autoencoder (AE) architecture. We propose a split architecture and use a pseudo-ground truth for abundances to guide the `unmixing network' (UN) optimization. Preceding the UN, an `approximation network' (AN) is proposed, which will improve the association between the centre pixel and its neighbourhood. Hence, it will accentuate spatial correlation in the abundances as its output is the input to the UN and the reference for the `mixing network' (MN). In the Guided Encoder-Decoder Architecture for Hyperspectral Unmixing with Spatial Smoothness (GAUSS), we proposed using one-hot encoded abundances as the pseudo-ground truth to guide the UN; computed using the k-means algorithm to exclude the use of prior HU methods. Furthermore, we release the single-layer constraint on MN by introducing the UN generated abundances in contrast to the standard AE for HU. Secondly, we experimented with two modifications on the pre-trained network using the GAUSS method. In GAUSS$_\textit{blind}$, we have concatenated the UN and the MN to back-propagate the reconstruction error gradients to the encoder. Then, in the GAUSS$_\textit{prime}$, abundance results of a signal processing (SP) method with reliable abundance results were used as the pseudo-ground truth with the GAUSS architecture. According to quantitative and graphical results for four experimental datasets, the three architectures either transcended or equated the performance of existing HU algorithms from both DL and SP domains.

comment: 16 pages, 6 figures

♻ ☆ EPNet++: Cascade Bi-directional Fusion for Multi-Modal 3D Object Detection

Zhe Liu, Tengteng Huang, Bingling Li, Xiwu Chen, Xi Wang, Xiang Bai

Recently, fusing the LiDAR point cloud and camera image to improve the performance and robustness of 3D object detection has received more and more attention, as these two modalities naturally possess strong complementarity. In this paper, we propose EPNet++ for multi-modal 3D object detection by introducing a novel Cascade Bi-directional Fusion~(CB-Fusion) module and a Multi-Modal Consistency~(MC) loss. More concretely, the proposed CB-Fusion module enhances point features with plentiful semantic information absorbed from the image features in a cascade bi-directional interaction fusion manner, leading to more powerful and discriminative feature representations. The MC loss explicitly guarantees the consistency between predicted scores from two modalities to obtain more comprehensive and reliable confidence scores. The experimental results on the KITTI, JRDB and SUN-RGBD datasets demonstrate the superiority of EPNet++ over the state-of-the-art methods. Besides, we emphasize a critical but easily overlooked problem, which is to explore the performance and robustness of a 3D detector in a sparser scene. Extensive experiments present that EPNet++ outperforms the existing SOTA methods with remarkable margins in highly sparse point cloud cases, which might be an available direction to reduce the expensive cost of LiDAR sensors. Code is available at: https://github.com/happinesslz/EPNetV2.

comment: Accepted by TPAMI-2022

♻ ☆ Self-Supervised 3D Traversability Estimation with Proxy Bank Guidance

Jihwan Bae, Junwon Seo, Taekyung Kim, Hae-gon Jeon, Kiho Kwak, Inwook Shim

Traversability estimation for mobile robots in off-road environments requires more than conventional semantic segmentation used in constrained environments like on-road conditions. Recently, approaches to learning a traversability estimation from past driving experiences in a self-supervised manner are arising as they can significantly reduce human labeling costs and labeling errors. However, the self-supervised data only provide supervision for the actually traversed regions, inducing epistemic uncertainty according to the scarcity of negative information. Negative data are rarely harvested as the system can be severely damaged while logging the data. To mitigate the uncertainty, we introduce a deep metric learning-based method to incorporate unlabeled data with a few positive and negative prototypes in order to leverage the uncertainty, which jointly learns using semantic segmentation and traversability regression. To firmly evaluate the proposed framework, we introduce a new evaluation metric that comprehensively evaluates the segmentation and regression. Additionally, we construct a driving dataset `Dtrail' in off-road environments with a mobile robot platform, which is composed of a wide variety of negative data. We examine our method on Dtrail as well as the publicly available SemanticKITTI dataset.

♻ ☆ EGSDE: Unpaired Image-to-Image Translation via Energy-Guided Stochastic Differential Equations NIPS 2022

Min Zhao, Fan Bao, Chongxuan Li, Jun Zhu

Score-based diffusion models (SBDMs) have achieved the SOTA FID results in unpaired image-to-image translation (I2I). However, we notice that existing methods totally ignore the training data in the source domain, leading to sub-optimal solutions for unpaired I2I. To this end, we propose energy-guided stochastic differential equations (EGSDE) that employs an energy function pretrained on both the source and target domains to guide the inference process of a pretrained SDE for realistic and faithful unpaired I2I. Building upon two feature extractors, we carefully design the energy function such that it encourages the transferred image to preserve the domain-independent features and discard domain-specific ones. Further, we provide an alternative explanation of the EGSDE as a product of experts, where each of the three experts (corresponding to the SDE and two feature extractors) solely contributes to faithfulness or realism. Empirically, we compare EGSDE to a large family of baselines on three widely-adopted unpaired I2I tasks under four metrics. EGSDE not only consistently outperforms existing SBDMs-based methods in almost all settings but also achieves the SOTA realism results without harming the faithful performance. Furthermore, EGSDE allows for flexible trade-offs between realism and faithfulness and we improve the realism results further (e.g., FID of 51.04 in Cat to Dog and FID of 50.43 in Wild to Dog on AFHQ) by tuning hyper-parameters. The code is available at https://github.com/ML-GSAI/EGSDE.

comment: NIPS 2022

♻ ☆ ScanNeRF: a Scalable Benchmark for Neural Radiance Fields WACV 2023

Luca De Luigi, Damiano Bolognini, Federico Domeniconi, Daniele De Gregorio, Matteo Poggi, Luigi Di Stefano

In this paper, we propose the first-ever real benchmark thought for evaluating Neural Radiance Fields (NeRFs) and, in general, Neural Rendering (NR) frameworks. We design and implement an effective pipeline for scanning real objects in quantity and effortlessly. Our scan station is built with less than 500$ hardware budget and can collect roughly 4000 images of a scanned object in just 5 minutes. Such a platform is used to build ScanNeRF, a dataset characterized by several train/val/test splits aimed at benchmarking the performance of modern NeRF methods under different conditions. Accordingly, we evaluate three cutting-edge NeRF variants on it to highlight their strengths and weaknesses. The dataset is available on our project page, together with an online benchmark to foster the development of better and better NeRFs.

comment: WACV 2023. The first three authors contributed equally. Project page: https://eyecan-ai.github.io/scannerf/

♻ ☆ Landslide4Sense: Reference Benchmark Data and Deep Learning Models for Landslide Detection

Omid Ghorbanzadeh, Yonghao Xu, Pedram Ghamisi, Michael Kopp, David Kreil

This study introduces \textit{Landslide4Sense}, a reference benchmark for landslide detection from remote sensing. The repository features 3,799 image patches fusing optical layers from Sentinel-2 sensors with the digital elevation model and slope layer derived from ALOS PALSAR. The added topographical information facilitates the accurate detection of landslide borders, which recent researches have shown to be challenging using optical data alone. The extensive data set supports deep learning (DL) studies in landslide detection and the development and validation of methods for the systematic update of landslide inventories. The benchmark data set has been collected at four different times and geographical locations: Iburi (September 2018), Kodagu (August 2018), Gorkha (April 2015), and Taiwan (August 2009). Each image pixel is labelled as belonging to a landslide or not, incorporating various sources and thorough manual annotation. We then evaluate the landslide detection performance of 11 state-of-the-art DL segmentation models: U-Net, ResU-Net, PSPNet, ContextNet, DeepLab-v2, DeepLab-v3+, FCN-8s, LinkNet, FRRN-A, FRRN-B, and SQNet. All models were trained from scratch on patches from one quarter of each study area and tested on independent patches from the other three quarters. Our experiments demonstrate that ResU-Net outperformed the other models for the landslide detection task. We make the multi-source landslide benchmark data (Landslide4Sense) and the tested DL models publicly available at \url{https://www.iarai.ac.at/landslide4sense}, establishing an important resource for remote sensing, computer vision, and machine learning communities in studies of image classification in general and applications to landslide detection in particular.

♻ ☆ Point Primitive Transformer for Long-Term 4D Point Cloud Video Understanding

Hao Wen, Yunze Liu, Jingwei Huang, Bo Duan, Li Yi

This paper proposes a 4D backbone for long-term point cloud video understanding. A typical way to capture spatial-temporal context is using 4Dconv or transformer without hierarchy. However, those methods are neither effective nor efficient enough due to camera motion, scene changes, sampling patterns, and the complexity of 4D data. To address those issues, we leverage the primitive plane as a mid-level representation to capture the long-term spatial-temporal context in 4D point cloud videos and propose a novel hierarchical backbone named Point Primitive Transformer(PPTr), which is mainly composed of intra-primitive point transformers and primitive transformers. Extensive experiments show that PPTr outperforms the previous state of the arts on different tasks.

♻ ☆ Target Aware Poisson-Gaussian Noise Parameters Estimation from Noisy Images

Étienne Objois, Kaan Okumuş, Nicolas Bähler

Digital sensors can lead to noisy results under many circumstances. To be able to remove the undesired noise from images, proper noise modeling and an accurate noise parameter estimation is crucial. In this project, we use a Poisson-Gaussian noise model for the raw-images captured by the sensor, as it fits the physical characteristics of the sensor closely. Moreover, we limit ourselves to the case where observed (noisy), and ground-truth (noise-free) image pairs are available. Using such pairs is beneficial for the noise estimation and is not widely studied in literature. Based on this model, we derive the theoretical maximum likelihood solution, discuss its practical implementation and optimization. Further, we propose two algorithms based on variance and cumulant statistics. Finally, we compare the results of our methods with two different approaches, a CNN we trained ourselves, and another one taken from literature. The comparison between all these methods shows that our algorithms outperform the others in terms of MSE and have good additional properties.

comment: 11 pages, 14 figures and 4 tables

♻ ☆ Combining Metric Learning and Attention Heads For Accurate and Efficient Multilabel Image Classification

Kirill Prokofiev, Vladislav Sovrasov

Multi-label image classification allows predicting a set of labels from a given image. Unlike multiclass classification, where only one label per image is assigned, such a setup is applicable for a broader range of applications. In this work we revisit two popular approaches to multilabel classification: transformer-based heads and labels relations information graph processing branches. Although transformer-based heads are considered to achieve better results than graph-based branches, we argue that with the proper training strategy, graph-based methods can demonstrate just a small accuracy drop, while spending less computational resources on inference. In our training strategy, instead of Asymmetric Loss (ASL), which is the de-facto standard for multilabel classification, we introduce its metric learning modification. In each binary classification sub-problem it operates with $L_2$ normalized feature vectors coming from a backbone and enforces angles between the normalized representations of positive and negative samples to be as large as possible. This results in providing a better discrimination ability, than binary cross entropy loss does on unnormalized features. With the proposed loss and training strategy, we obtain SOTA results among single modality methods on widespread multilabel classification benchmarks such as MS-COCO, PASCAL-VOC, NUS-Wide and Visual Genome 500. Source code of our method is available as a part of the OpenVINO Training Extensions https://github.com/openvinotoolkit/deep-object-reid/tree/multilabel

comment: Accepted at VISAPP 2023

♻ ☆ Diffusion Models in Vision: A Survey

Florinel-Alin Croitoru, Vlad Hondru, Radu Tudor Ionescu, Mubarak Shah

Denoising diffusion models represent a recent emerging topic in computer vision, demonstrating remarkable results in the area of generative modeling. A diffusion model is a deep generative model that is based on two stages, a forward diffusion stage and a reverse diffusion stage. In the forward diffusion stage, the input data is gradually perturbed over several steps by adding Gaussian noise. In the reverse stage, a model is tasked at recovering the original input data by learning to gradually reverse the diffusion process, step by step. Diffusion models are widely appreciated for the quality and diversity of the generated samples, despite their known computational burdens, i.e. low speeds due to the high number of steps involved during sampling. In this survey, we provide a comprehensive review of articles on denoising diffusion models applied in vision, comprising both theoretical and practical contributions in the field. First, we identify and present three generic diffusion modeling frameworks, which are based on denoising diffusion probabilistic models, noise conditioned score networks, and stochastic differential equations. We further discuss the relations between diffusion models and other deep generative models, including variational auto-encoders, generative adversarial networks, energy-based models, autoregressive models and normalizing flows. Then, we introduce a multi-perspective categorization of diffusion models applied in computer vision. Finally, we illustrate the current limitations of diffusion models and envision some interesting directions for future research.

comment: 24 pages, 3 figures

♻ ☆ Unbiased 4D: Monocular 4D Reconstruction with a Neural Deformation Model

Erik C. M. Johnson, Marc Habermann, Soshi Shimada, Vladislav Golyanik, Christian Theobalt

Capturing general deforming scenes from monocular RGB video is crucial for many computer graphics and vision applications. However, current approaches suffer from drawbacks such as struggling with large scene deformations, inaccurate shape completion or requiring 2D point tracks. In contrast, our method, Ub4D, handles large deformations, performs shape completion in occluded regions, and can operate on monocular RGB videos directly by using differentiable volume rendering. This technique includes three new in the context of non-rigid 3D reconstruction components, i.e., 1) A coordinate-based and implicit neural representation for non-rigid scenes, which in conjunction with differentiable volume rendering enables an unbiased reconstruction of dynamic scenes, 2) a proof that extends the unbiased formulation of volume rendering to dynamic scenes, and 3) a novel dynamic scene flow loss, which enables the reconstruction of larger deformations by leveraging the coarse estimates of other methods. Results on our new dataset, which will be made publicly available, demonstrate a clear improvement over the state of the art in terms of surface reconstruction accuracy and robustness to large deformations.

comment: 26 pages, 17 figures, 8 tables

♻ ☆ Complete-to-Partial 4D Distillation for Self-Supervised Point Cloud Sequence Representation Learning

Zhuoyang Zhang, Yuhao Dong, Yunze Liu, Li Yi

Recent work on 4D point cloud sequences has attracted a lot of attention. However, obtaining exhaustively labeled 4D datasets is often very expensive and laborious, so it is especially important to investigate how to utilize raw unlabeled data. However, most existing self-supervised point cloud representation learning methods only consider geometry from a static snapshot omitting the fact that sequential observations of dynamic scenes could reveal more comprehensive geometric details. And the video representation learning frameworks mostly model motion as image space flows, let alone being 3D-geometric-aware. To overcome such issues, this paper proposes a new 4D self-supervised pre-training method called Complete-to-Partial 4D Distillation. Our key idea is to formulate 4D self-supervised representation learning as a teacher-student knowledge distillation framework and let the student learn useful 4D representations with the guidance of the teacher. Experiments show that this approach significantly outperforms previous pre-training approaches on a wide range of 4D point cloud sequence understanding tasks including indoor and outdoor scenarios.

♻ ☆ OG-SGG: Ontology-Guided Scene Graph Generation. A Case Study in Transfer Learning for Telepresence Robotics

Fernando Amodeo, Fernando Caballero, Natalia Díaz-Rodríguez, Luis Merino

Scene graph generation from images is a task of great interest to applications such as robotics, because graphs are the main way to represent knowledge about the world and regulate human-robot interactions in tasks such as Visual Question Answering (VQA). Unfortunately, its corresponding area of machine learning is still relatively in its infancy, and the solutions currently offered do not specialize well in concrete usage scenarios. Specifically, they do not take existing "expert" knowledge about the domain world into account; and that might indeed be necessary in order to provide the level of reliability demanded by the use case scenarios. In this paper, we propose an initial approximation to a framework called Ontology-Guided Scene Graph Generation (OG-SGG), that can improve the performance of an existing machine learning based scene graph generator using prior knowledge supplied in the form of an ontology (specifically, using the axioms defined within); and we present results evaluated on a specific scenario founded in telepresence robotics. These results show quantitative and qualitative improvements in the generated scene graphs.

comment: 20 pages; version accepted and published in IEEE Access

♻ ☆ Painterly Image Harmonization in Dual Domains AAAI2023

Junyan Cao, Yan Hong, Li Niu

Image harmonization aims to produce visually harmonious composite images by adjusting the foreground appearance to be compatible with the background. When the composite image has photographic foreground and painterly background, the task is called painterly image harmonization. There are only few works on this task, which are either time-consuming or weak in generating well-harmonized results. In this work, we propose a novel painterly harmonization network consisting of a dual-domain generator and a dual-domain discriminator, which harmonizes the composite image in both spatial domain and frequency domain. The dual-domain generator performs harmonization by using AdaIn modules in the spatial domain and our proposed ResFFT modules in the frequency domain. The dual-domain discriminator attempts to distinguish the inharmonious patches based on the spatial feature and frequency feature of each patch, which can enhance the ability of generator in an adversarial manner. Extensive experiments on the benchmark dataset show the effectiveness of our method. Our code and model are available at https://github.com/bcmi/PHDNet-Painterly-Image-Harmonization.

comment: Accepted by AAAI2023

♻ ☆ A Comprehensive Survey and Taxonomy on Single Image Dehazing Based on Deep Learning

Jie Gui, Xiaofeng Cong, Yuan Cao, Wenqi Ren, Jun Zhang, Jing Zhang, Jiuxin Cao, Dacheng Tao

With the development of convolutional neural networks, hundreds of deep learning based dehazing methods have been proposed. In this paper, we provide a comprehensive survey on supervised, semi-supervised, and unsupervised single image dehazing. We first discuss the physical model, datasets, network modules, loss functions, and evaluation metrics that are commonly used. Then, the main contributions of various dehazing algorithms are categorized and summarized. Further, quantitative and qualitative experiments of various baseline methods are carried out. Finally, the unsolved issues and challenges that can inspire the future research are pointed out. A collection of useful dehazing materials is available at \url{https://github.com/Xiaofeng-life/AwesomeDehazing}.

comment: This paper is accepted by ACM Computing Surveys

♻ ☆ Towards Practical Verification of Machine Learning: The Case of Computer Vision Systems

Kexin Pei, Linjie Zhu, Yinzhi Cao, Junfeng Yang, Carl Vondrick, Suman Jana

Due to the increasing usage of machine learning (ML) techniques in security- and safety-critical domains, such as autonomous systems and medical diagnosis, ensuring correct behavior of ML systems, especially for different corner cases, is of growing importance. In this paper, we propose a generic framework for evaluating security and robustness of ML systems using different real-world safety properties. We further design, implement and evaluate VeriVis, a scalable methodology that can verify a diverse set of safety properties for state-of-the-art computer vision systems with only blackbox access. VeriVis leverage different input space reduction techniques for efficient verification of different safety properties. VeriVis is able to find thousands of safety violations in fifteen state-of-the-art computer vision systems including ten Deep Neural Networks (DNNs) such as Inception-v3 and Nvidia's Dave self-driving system with thousands of neurons as well as five commercial third-party vision APIs including Google vision and Clarifai for twelve different safety properties. Furthermore, VeriVis can successfully verify local safety properties, on average, for around 31.7% of the test images. VeriVis finds up to 64.8x more violations than existing gradient-based methods that, unlike VeriVis, cannot ensure non-existence of any violations. Finally, we show that retraining using the safety violations detected by VeriVis can reduce the average number of violations up to 60.2%.

♻ ☆ Improvised Aerial Object Detection approach for YOLOv3 Using Weighted Luminance

Sai Ganesh CS, Aouthithiye Barathwaj SR Y, R. Swethaa S, R. Azhagumurugan

Aerial imaging plays a crucial role in navigation and data acquisition for unmanned aerial vehicles and satellite imaging systems. In recent days, the employment of drones has been escalated in several applications that are not limited to surveillance, delivery systems, aerial warfare, and agricultural activities. Aerial imaging of ground targets is highly challenging because of various factors that affect light propagation through different mediums. Several convolutional neural network-based object detection algorithms that are developed require more robustness when applied in the field of aerial imaging and remote sensing. In order to handle the adverse effects of light propagation with respect to time and solar radiance, adaptive RGB filters for grayscale imaging based on weighted luminance are introduced that extensively solve the problem of rayleigh scattering effect. Images of objects that are easily diminished by rayleigh scattering are acquired in various timezones. The acquired images are labelled precisely and subjected to training and validation. The results show that the proposed method detects the object more accurately and efficiently than the traditional YOLOv3 approach.

comment: 17 pages, 4 figures, Journal Expert Systems with Applications

♻ ☆ Learning from Training Dynamics: Identifying Mislabeled Data Beyond Manually Designed Features AAAI23

Qingrui Jia, Xuhong Li, Lei Yu, Jiang Bian, Penghao Zhao, Shupeng Li, Haoyi Xiong, Dejing Dou

While mislabeled or ambiguously-labeled samples in the training set could negatively affect the performance of deep models, diagnosing the dataset and identifying mislabeled samples helps to improve the generalization power. Training dynamics, i.e., the traces left by iterations of optimization algorithms, have recently been proved to be effective to localize mislabeled samples with hand-crafted features. In this paper, beyond manually designed features, we introduce a novel learning-based solution, leveraging a noise detector, instanced by an LSTM network, which learns to predict whether a sample was mislabeled using the raw training dynamics as input. Specifically, the proposed method trains the noise detector in a supervised manner using the dataset with synthesized label noises and can adapt to various datasets (either naturally or synthesized label-noised) without retraining. We conduct extensive experiments to evaluate the proposed method. We train the noise detector based on the synthesized label-noised CIFAR dataset and test such noise detector on Tiny ImageNet, CUB-200, Caltech-256, WebVision and Clothing1M. Results show that the proposed method precisely detects mislabeled samples on various datasets without further adaptation, and outperforms state-of-the-art methods. Besides, more experiments demonstrate that the mislabel identification can guide a label correction, namely data debugging, providing orthogonal improvements of algorithm-centric state-of-the-art techniques from the data aspect.

comment: AAAI23 accepted Conference Paper

♻ ☆ Joint Feature Learning and Relation Modeling for Tracking: A One-Stream Framework ECCV 2022

Botao Ye, Hong Chang, Bingpeng Ma, Shiguang Shan, Xilin Chen

The current popular two-stream, two-stage tracking framework extracts the template and the search region features separately and then performs relation modeling, thus the extracted features lack the awareness of the target and have limited target-background discriminability. To tackle the above issue, we propose a novel one-stream tracking (OSTrack) framework that unifies feature learning and relation modeling by bridging the template-search image pairs with bidirectional information flows. In this way, discriminative target-oriented features can be dynamically extracted by mutual guidance. Since no extra heavy relation modeling module is needed and the implementation is highly parallelized, the proposed tracker runs at a fast speed. To further improve the inference efficiency, an in-network candidate early elimination module is proposed based on the strong similarity prior calculated in the one-stream framework. As a unified framework, OSTrack achieves state-of-the-art performance on multiple benchmarks, in particular, it shows impressive results on the one-shot tracking benchmark GOT-10k, i.e., achieving 73.7% AO, improving the existing best result (SwinTrack) by 4.3\%. Besides, our method maintains a good performance-speed trade-off and shows faster convergence. The code and models are available at https://github.com/botaoye/OSTrack.

comment: Accepted by ECCV 2022

♻ ☆ Texts as Images in Prompt Tuning for Multi-Label Image Recognition

Zixian Guo, Bowen Dong, Zhilong Ji, Jinfeng Bai, Yiwen Guo, Wangmeng Zuo

Prompt tuning has been employed as an efficient way to adapt large vision-language pre-trained models (e.g. CLIP) to various downstream tasks in data-limited or label-limited settings. Nonetheless, visual data (e.g., images) is by default prerequisite for learning prompts in existing methods. In this work, we advocate that the effectiveness of image-text contrastive learning in aligning the two modalities (for training CLIP) further makes it feasible to treat texts as images for prompt tuning and introduce TaI prompting. In contrast to the visual data, text descriptions are easy to collect, and their class labels can be directly derived. Particularly, we apply TaI prompting to multi-label image recognition, where sentences in the wild serve as alternatives to images for prompt tuning. Moreover, with TaI, double-grained prompt tuning (TaI-DPT) is further presented to extract both coarse-grained and fine-grained embeddings for enhancing the multi-label recognition performance. Experimental results show that our proposed TaI-DPT outperforms zero-shot CLIP by a large margin on multiple benchmarks, e.g., MS-COCO, VOC2007, and NUS-WIDE, while it can be combined with existing methods of prompting from images to improve recognition performance further. Code is released at https://github.com/guozix/TaI-DPT.

♻ ☆ Method to Classify Skin Lesions using Dermoscopic images

Dusa Sai Charan, Hemanth Nadipineni, Subin Sahayam, Umarani Jayaraman

Skin cancer is the most common cancer in the existing world constituting one-third of the cancer cases. Benign skin cancers are not fatal, can be cured with proper medication. But it is not the same as the malignant skin cancers. In the case of malignant melanoma, in its peak stage, the maximum life expectancy is less than or equal to 5 years. But, it can be cured if detected in early stages. Though there are numerous clinical procedures, the accuracy of diagnosis falls between 49% to 81% and is time-consuming. So, dermoscopy has been brought into the picture. It helped in increasing the accuracy of diagnosis but could not demolish the error-prone behaviour. A quick and less error-prone solution is needed to diagnose this majorly growing skin cancer. This project deals with the usage of deep learning in skin lesion classification. In this project, an automated model for skin lesion classification using dermoscopic images has been developed with CNN(Convolution Neural Networks) as a training model. Convolution neural networks are known for capturing features of an image. So, they are preferred in analyzing medical images to find the characteristics that drive the model towards success. Techniques like data augmentation for tackling class imbalance, segmentation for focusing on the region of interest and 10-fold cross-validation to make the model robust have been brought into the picture. This project also includes usage of certain preprocessing techniques like brightening the images using piece-wise linear transformation function, grayscale conversion of the image, resize the image. This project throws a set of valuable insights on how the accuracy of the model hikes with the bringing of new input strategies, preprocessing techniques. The best accuracy this model could achieve is 0.886.

comment: 16 pages, 14 figures

♻ ☆ Correspondence Distillation from NeRF-based GAN

Yushi Lan, Chen Change Loy, Bo Dai

The neural radiance field (NeRF) has shown promising results in preserving the fine details of objects and scenes. However, unlike mesh-based representations, it remains an open problem to build dense correspondences across different NeRFs of the same category, which is essential in many downstream tasks. The main difficulties of this problem lie in the implicit nature of NeRF and the lack of ground-truth correspondence annotations. In this paper, we show it is possible to bypass these challenges by leveraging the rich semantics and structural priors encapsulated in a pre-trained NeRF-based GAN. Specifically, we exploit such priors from three aspects, namely 1) a dual deformation field that takes latent codes as global structural indicators, 2) a learning objective that regards generator features as geometric-aware local descriptors, and 3) a source of infinite object-specific NeRF samples. Our experiments demonstrate that such priors lead to 3D dense correspondence that is accurate, smooth, and robust. We also show that established dense correspondence across NeRFs can effectively enable many NeRF-based downstream applications such as texture transfer.

comment: Project page: https://nirvanalan.github.io/projects/DDF/index.html

♻ ☆ Delving into the Openness of CLIP

Shuhuai Ren, Lei Li, Xuancheng Ren, Guangxiang Zhao, Xu Sun

Contrastive Language-Image Pre-training (CLIP) has demonstrated great potential in realizing open-vocabulary visual recognition in a matching style, due to its holistic use of natural language supervision that covers unconstrained real-world visual concepts. However, it is, in turn, also difficult to evaluate and analyze the openness of CLIP-like models, since they are in theory open to any vocabulary but the actual accuracy varies. To address the insufficiency of conventional studies on openness, we resort to an incremental perspective and define the extensibility, which essentially approximates the model's ability to deal with new visual concepts, by evaluating openness through vocabulary expansions. Our evaluation based on extensibility shows that CLIP-like models are hardly truly open and their performance degrades as the vocabulary expands to different degrees. Further analysis reveals that the over-estimation of openness is not because CLIP-like models fail to capture the general similarity of image and text features of novel visual concepts, but because of the confusion among competing text features, that is, they are not stable with respect to the vocabulary. In light of this, we propose to improve the openness of CLIP in the feature space by enforcing the distinguishability of text features. Our method retrieves relevant texts from the pre-training corpus to enhance prompts for inference, which boosts the extensibility and stability of CLIP even without fine-tuning.

comment: 22 pages, 12 figures. Code is available at https://github.com/lancopku/clip-openness

♻ ☆ Stereo Image Rain Removal via Dual-View Mutual Attention

Yanyan Wei, Zhao Zhang, Zhongqiu Zhao, Yang Zhao, Richang Hong, Yi Yang

Stereo images, containing left and right view images with disparity, are utilized in solving low-vision tasks recently, e.g., rain removal and super-resolution. Stereo image restoration methods usually obtain better performance than monocular methods by learning the disparity between dual views either implicitly or explicitly. However, existing stereo rain removal methods still cannot make full use of the complementary information between two views, and we find it is because: 1) the rain streaks have more complex distributions in directions and densities, which severely damage the complementary information and pose greater challenges; 2) the disparity estimation is not accurate enough due to the imperfect fusion mechanism for the features between two views. To overcome such limitations, we propose a new \underline{Stereo} \underline{I}mage \underline{R}ain \underline{R}emoval method (StereoIRR) via sufficient interaction between two views, which incorporates: 1) a new Dual-view Mutual Attention (DMA) mechanism which generates mutual attention maps by taking left and right views as key information for each other to facilitate cross-view feature fusion; 2) a long-range and cross-view interaction, which is constructed with basic blocks and dual-view mutual attention, can alleviate the adverse effect of rain on complementary information to help the features of stereo images to get long-range and cross-view interaction and fusion. Notably, StereoIRR outperforms other related monocular and stereo image rain removal methods on several datasets. Our codes and datasets will be released.

♻ ☆ DUET: A Tuning-Free Device-Cloud Collaborative Parameters Generation Framework for Efficient Device Model Generalization

Zheqi Lv, Wenqiao Zhang, Shengyu Zhang, Kun Kuang, Feng Wang, Yongwei Wang, Zhengyu Chen, Tao Shen, Hongxia Yang, Beng chin Ooi, Fei Wu

Device Model Generalization (DMG) is a practical yet under-investigated research topic for on-device machine learning applications. It aims to improve the generalization ability of pre-trained models when deployed on resource-constrained devices, such as improving the performance of pre-trained cloud models on smart mobiles. While quite a lot of works have investigated the data distribution shift across clouds and devices, most of them focus on model fine-tuning on personalized data for individual devices to facilitate DMG. Despite their promising, these approaches require on-device re-training, which is practically infeasible due to the overfitting problem and high time delay when performing gradient calculation on real-time data. In this paper, we argue that the computational cost brought by fine-tuning can be rather unnecessary. We consequently present a novel perspective to improving DMG without increasing computational cost, i.e., device-specific parameter generation which directly maps data distribution to parameters. Specifically, we propose an efficient Device-cloUd collaborative parametErs generaTion framework DUET. DUET is deployed on a powerful cloud server that only requires the low cost of forwarding propagation and low time delay of data transmission between the device and the cloud. By doing so, DUET can rehearse the device-specific model weight realizations conditioned on the personalized real-time data for an individual device. Importantly, our DUET elegantly connects the cloud and device as a 'duet' collaboration, frees the DMG from fine-tuning, and enables a faster and more accurate DMG paradigm. We conduct an extensive experimental study of DUET on three public datasets, and the experimental results confirm our framework's effectiveness and generalisability for different DMG tasks.

♻ ☆ Unifying Vision, Text, and Layout for Universal Document Processing

Zineng Tang, Ziyi Yang, Guoxin Wang, Yuwei Fang, Yang Liu, Chenguang Zhu, Michael Zeng, Cha Zhang, Mohit Bansal

We propose Universal Document Processing (UDOP), a foundation Document AI model which unifies text, image, and layout modalities together with varied task formats, including document understanding and generation. UDOP leverages the spatial correlation between textual content and document image to model image, text, and layout modalities with one uniform representation. With a novel Vision-Text-Layout Transformer, UDOP unifies pretraining and multi-domain downstream tasks into a prompt-based sequence generation scheme. UDOP is pretrained on both large-scale unlabeled document corpora using innovative self-supervised objectives and diverse labeled data. UDOP also learns to generate document images from text and layout modalities via masked image reconstruction. To the best of our knowledge, this is the first time in the field of document AI that one model simultaneously achieves high-quality neural document editing and content customization. Our method sets the state-of-the-art on 9 Document AI tasks, e.g., document understanding and QA, across diverse data domains like finance reports, academic papers, and websites. UDOP ranks first on the leaderboard of the Document Understanding Benchmark (DUE).

♻ ☆ Cross-view Geo-localization via Learning Disentangled Geometric Layout Correspondence

Xiaohan Zhang, Xingyu Li, Waqas Sultani, Yi Zhou, Safwan Wshah

Cross-view geo-localization aims to estimate the location of a query ground image by matching it to a reference geo-tagged aerial images database. As an extremely challenging task, its difficulties root in the drastic view changes and different capturing time between two views. Despite these difficulties, recent works achieve outstanding progress on cross-view geo-localization benchmarks. However, existing methods still suffer from poor performance on the cross-area benchmarks, in which the training and testing data are captured from two different regions. We attribute this deficiency to the lack of ability to extract the spatial configuration of visual feature layouts and models' overfitting on low-level details from the training set. In this paper, we propose GeoDTR which explicitly disentangles geometric information from raw features and learns the spatial correlations among visual features from aerial and ground pairs with a novel geometric layout extractor module. This module generates a set of geometric layout descriptors, modulating the raw features and producing high-quality latent representations. In addition, we elaborate on two categories of data augmentations, (i) Layout simulation, which varies the spatial configuration while keeping the low-level details intact. (ii) Semantic augmentation, which alters the low-level details and encourages the model to capture spatial configurations. These augmentations help to improve the performance of the cross-view geo-localization models, especially on the cross-area benchmarks. Moreover, we propose a counterfactual-based learning process to benefit the geometric layout extractor in exploring spatial information. Extensive experiments show that GeoDTR not only achieves state-of-the-art results but also significantly boosts the performance on same-area and cross-area benchmarks.

♻ ☆ Refine and Represent: Region-to-Object Representation Learning

Akash Gokul, Konstantinos Kallidromitis, Shufan Li, Yusuke Kato, Kazuki Kozuka, Trevor Darrell, Colorado J Reed

Recent works in self-supervised learning have demonstrated strong performance on scene-level dense prediction tasks by pretraining with object-centric or region-based correspondence objectives. In this paper, we present Region-to-Object Representation Learning (R2O) which unifies region-based and object-centric pretraining. R2O operates by training an encoder to dynamically refine region-based segments into object-centric masks and then jointly learns representations of the contents within the mask. R2O uses a "region refinement module" to group small image regions, generated using a region-level prior, into larger regions which tend to correspond to objects by clustering region-level features. As pretraining progresses, R2O follows a region-to-object curriculum which encourages learning region-level features early on and gradually progresses to train object-centric representations. Representations learned using R2O lead to state-of-the art performance in semantic segmentation for PASCAL VOC (+0.7 mIOU) and Cityscapes (+0.4 mIOU) and instance segmentation on MS COCO (+0.3 mask AP). Further, after pretraining on ImageNet, R2O pretrained models are able to surpass existing state-of-the-art in unsupervised object segmentation on the Caltech-UCSD Birds 200-2011 dataset (+2.9 mIoU) without any further training. We provide the code/models from this work at https://github.com/KKallidromitis/r2o.

♻ ☆ MammoDL: Mammographic Breast Density Estimation using Federated Learning

Ramya Muthukrishnan, Angelina Heyler, Keshava Katti, Sarthak Pati, Walter Mankowski, Aprupa Alahari, Michael Sanborn, Emily F. Conant, Christopher Scott, Stacey Winham, Celine Vachon, Pratik Chaudhari, Despina Kontos, Spyridon Bakas

Assessing breast cancer risk from imaging remains a subjective process, in which radiologists employ simple computer aided detection (CAD) systems or qualitative visual assessment to estimate breast percent density (PD). Machine learning (ML) models have become the most promising way to quantify breast cancer risk for early, accurate, and equitable diagnoses, but training such models in medical research is often restricted to small, single-institution data. Since patient demographics and imaging characteristics may vary considerably across imaging sites, models trained on single-institution data tend not to generalize well. In response to this problem, MammoDL is proposed, an open-source software tool that leverages a U-Net architecture to accurately estimate breast PD and complexity from mammography. With the Open Federated Learning (OpenFL) library, this solution enables secure training on datasets across multiple institutions. MammoDL is a leaner, more flexible model than its predecessors, boasting improved generalization due to federation-enabled training on larger, more representative datasets.

comment: Breast Cancer Risk, Digital Mammography, Breast Density, Deep Learning, Machine Learning, Federated Learning, OpenFL

♻ ☆ CLIP is Also an Efficient Segmenter: A Text-Driven Approach for Weakly Supervised Semantic Segmentation

Yuqi Lin, Minghao Chen, Wenxiao Wang, Boxi Wu, Ke Li, Binbin Lin, Haifeng Liu, Xiaofei He

Weakly supervised semantic segmentation (WSSS) with image-level labels is a challenging task in computer vision. Mainstream approaches follow a multi-stage framework and suffer from high training costs. In this paper, we explore the potential of Contrastive Language-Image Pre-training models (CLIP) to localize different categories with only image-level labels and without any further training. To efficiently generate high-quality segmentation masks from CLIP, we propose a novel framework called CLIP-ES for WSSS. Our framework improves all three stages of WSSS with special designs for CLIP: 1) We introduce the softmax function into GradCAM and exploit the zero-shot ability of CLIP to suppress the confusion caused by non-target classes and backgrounds. Meanwhile, to take full advantage of CLIP, we re-explore text inputs under the WSSS setting and customize two text-driven strategies: sharpness-based prompt selection and synonym fusion. 2) To simplify the stage of CAM refinement, we propose a real-time class-aware attention-based affinity (CAA) module based on the inherent multi-head self-attention (MHSA) in CLIP-ViTs. 3) When training the final segmentation model with the masks generated by CLIP, we introduced a confidence-guided loss (CGL) to mitigate noise and focus on confident regions. Our proposed framework dramatically reduces the cost of training for WSSS and shows the capability of localizing objects in CLIP. Our CLIP-ES achieves SOTA performance on Pascal VOC 2012 and MS COCO 2014 while only taking 10% time of previous methods for the pseudo mask generation. Code is available at https://github.com/linyq2117/CLIP-ES.

comment: 13 pages, 8 figures